ScienceScopeP U B L I C A T I O N O F T H E C S I R | V O L U M E 5 | N U M B E R 2 | M A Y 2 0 1 1S O U T H A F R I C A ' S C O U N C I L F O R S C I E N T I F I C A N D I N D U S T R I A L R E S E A R C H

www.csir.co.za

Information and communications technology

THE CSIR’S OPERATING UNITS, NATIONAL RESEARCH CENTRES AND SERVICES

> CSIR Biosciences Pretoria 012 841-3260 Modderfontein 011 605-2700

> CSIR Built Environment Pretoria 012 841-3871 Stellenbosch 021 888-2508

> CSIR Defence, Peace, Safety and Security Pretoria 012 841-2780

> CSIR Materials Science and Manufacturing Pretoria 012 841-4392 Johannesburg 011 482-1300 Port Elizabeth 041 508-3200 Cape Town 021 685-4329

> CSIR Modelling and Digital Science Pretoria 012 841-3298

> CSIR Natural Resources and the Environment Pretoria 012 841-4005 Stellenbosch 021 888-2400 Durban 031 242-2300 Pietermaritzburg 033 260-5446 Nelspruit 013 759-8036

> CSIR Consulting and Analytical Services Pretoria 012 841-2525 Stellenbosch 021 658-2766 Cottesloe 011 482-1300 Modderfontein 011 605-2452

> CSIR Meraka Institute Pretoria 012 841-3028 Cape Town (Centre for High Performance Computing) 021 658-2740

> CSIR National Laser Centre Pretoria 012 841-4188

> CSIR Centre for Mining Innovation Johannesburg 011 358-0000

As of 1 April 2011, the CSIR Satellite Applications Centre forms part of the South African National Space Agency (SANSA)

Printed on Sappi Triple GreenTM paper and using an environmentally-responsible manufacturing process. Sugar cane fibre (bagasse) is used – a fast growing, annually renewable resource. All pulp is bleached using an elemental chlorine-free process, complying with stringent regulations of the US Environmental Protection Agency and the World Health Organization. Pulping chemicals go through a chemical recovery plant, waste water undergoes rigorous treatment and solid waste is recovered to use as soil enhancer to regenerate harvestedsugar cane fibre lands.

Connections extraordinary everywhere: ‘Internet of Things’ 6

ICT research, development and innovation: Absolutely relevant to earth observation 8

From satellites to pathogens – building a global system of systems to understand planet Earth 10

From satellites to pathogens – mapping the environmental drivers of cholera epidemics on a larger scale 12

Enhancing African environmental management through ICT 14

Scientific workflows: Keeping track of useful earth observation science 15

High performance computing revolutionises earth observation image processing 16

New systems developed to study water quality and eutrophication 17

Research aids sustainable planning of cities 30 years ahead 18

Pioneering the spoken web in South African English, isiZulu and Afrikaans 20

Uncovering new knowledge in scientific domains using ontologies 21

Empowering communities to build networks 22

Information services to support entrepreneurial development 24

Safeguarding the future of indigenous knowledge through ICT? 25

Making the most of infrastructure: The role of the SAGrid 27

An African e-infrastructure in the making 28

Unlocking the potential of research networking 30

SAFIPA’s curtain drawing 32

Reaping competitive advantage through high performance computing and data curation 33

Extra sensory perception 35

Community internet television – live broadcasting for the masses 36

mLabs: Empowering through locally-developed mobile applications 39

ICT in the service of emerging economies 40

Novel web-based calculator to estimate emissions and general carbon footprint 43

INFOSCOPE

RECAPPING

3

6

27

33

44

INTRODUCTION 2

ICT AND SOCIET Y

ICT AND HUMAN CAPITAL DEVELOPMENT

ICT AND THE ECONOMY

Compiled by CSIR Strategic Communications and Stakeholder Relations

Editorial team:Thabo Ditsele (Editor & Project Manager)Collette Vosloo (Content Manager)

WritersBandile SikwaneBiffy van RooyenBonang TselaneHilda van RooyenMzimasi GcukumanaPetro LowiesWiida Basson

ContributorsLee AnnamalaiDr Konrad WesselsDr Bob Scholes

PhotographyLesa van RooyenPieter UysLoretta Steynistock.com

Cover illustration:by Loretta Steyn from images by Dreamstime and istock

Design and production: Loretta Steyn Graphic Design Studio

| 2 |

In 1948, Claude Shannon, a young engineer and mathematician, published ‘A Mathematical Theoryof Communication’. With this landmark paper, Shannon laid the foundation for information theory.

ON 30 JUNE in the same year,Bell Labs unveiled the invention ofthe transistor by Bardeen andBrattain, discovered almost acci-dentally during experimentsaimed at understanding the be-haviour of electrons at the inter-face between a metal and asemiconductor and why the ‘fieldeffect’ semiconductor amplifierproposed by Shokley did notwork.

In the 1956 award ceremonyspeech for the Nobel Prize inPhysics, their achievement wasdescribed thus: “Doctor Shockley,Doctor Bardeen and Doctor Brat-tain. The summit of Everest wasreached by a small party of ardentclimbers. Working from an ad-vance base, they succeeded. Morethan a generation of mountaineershad toiled to establish that base.Your assault on the semiconductorproblem was likewise launchedfrom a high-altitude camp, con-tributed by many scientists. Yours,too, was a supreme effort – of fore-sight, ingenuity and perseverance,exercised individually and as ateam. Surely, supreme joy befallsthe man to whom those breath-taking vistas from the summit unfold. You must have felt it, overwhelmingly. This joy is nowshared by those who laboured atthe base. Shared, too, is the chal-lenge of untrodden territory, nowseen for the first time, calling for anew scientific attack.”

Shannon's work on informationtheory has subsequently had sig-nificant implication in engineering

and computer science. The inven-tion of the transistor eventually ledto digital electronics enabling themodern field of information andcommunication technology or ICT from the convergence of thepreviously separate fields oftelecommunication and informa-tion technology. Other milestoneson the road to the now nearlyubiquitous ICT, from twitter andfacebook revolutions to iPads, include the development of inte-grated circuit technology (JackKilby, 1958), Ethernet (Robert Metcalfe, 1976), the Internet protocol (Vint Cerf and Bob Kahn,1974), optic fibre (Charles KuenKao and other between 1966 and1975 when the first non-experi-mental fibre-optic link was installed by the Dorset, UK police),GSM mobile digital telephony (first operated by Radiolinjain inFinland on 19 September, 1988)and the World Wide Web (TimBerners Lee, 1989). Two other keydevelopments that are enablingthe digital information revolutionare cheap networked sensors andglobal positioning systems.

Today, the impact of ICT on eco-nomic development and improvedquality of life is well established.Studies by the World Bank indi-cate that in low- and middle-income countries, every 10 per-centage point increase in broad-band penetration accelerateseconomic growth by 1,38 percent-age points. In South Africa, the Financial Mail, in a feature articleon broadband quoted the GSM Association, stated, “Wireless

broadband and related industriescould generate about 28 000 newjobs and 1.8% GDP by 2015.”

CSIR's work in ICT research anddevelopment takes its cue fromthe concept of the information society. Our vision is to contributeto South Africa's development towards and advanced informa-tion society in which everyone canbenefit from their enhanced abilityto create, manipulate, organize,transmit, store, and act on infor-mation.

This edition of ScienceScope high-lights a number of areas of ICT research in the CSIR. It shows howour work contributes to society,human capital development, andthe economy.

Advances in ICT, brought about by research and experimental development, are worthy of pursuit both in terms of enablingimpact in areas such as health, energy and the environment, and as an area of impact in itself.Hence we are taking up the ‘challenge of untrodden territory’by investing our resources in aportfolio of initiatives aimed at scientific and technological break-throughs that will broaden andenhance access to ICT and enablenovel uses of advanced ICT.

Dr Sibusiso Sibisi,CSIR President and CEO

I N T R O D U C T I O N

infoscopeNews snippets from around the CSIR

THE UNPRECEDENTED increase in the number and size of potholesfollowing the past summer rainfallseasons has seen a public outcryand politicians ring-fencing fund-ing to address this situation.

The fixing of potholes is the re-sponsibility of road owners. SinceFebruary 2011, CSIR researchershave been training scores of roadinspection and maintenanceteams nationally to ensure appro-priate and effective fixing of potholes. The national Depart-ment of Public Works already sentmore than 300 people for thetraining to date, with the SouthAfrican Road Federation (SARF)and the Asphalt Academy part-nering with the CSIR to presenttraining sessions nationally. Metros are starting to take up theopportunity for training, with theCity of Tshwane being the first.

The CSIR published appropriateresearch and development (R&D)solutions for the pothole problemin a technical guideline in Decem-ber 2010. The guide – used duringthe training – deals with thecauses of potholes, their identifica-

tion and the various repair methods for different categories of potholes. The CSIR also pub-lished a shorter, non-technicaldocument (both are freely avai-lable at – www.csir.co.za /pothole_guides).

The guide is the only one of itskind in South Africa and includesan innovative pothole classifica-tion system developed by theCSIR. It is used to classify potholesinto one of seven categories, witheach category requiring a differentrepair technique to address thefundamental cause of the potholeduring the repair process.

– Hilda van Rooyen

CSIR assists with pothole problem through R&D solution

Economic growthhampered by logistics costsSOUTH AFRICA’S consistently high cost of logistics decreased by just 1.2% between2008 and 2009, from 14.7% to 13.5% ofGross Domestic Product (GDP). Within a recessionary context, this points to an un-derperforming logistics sector, according tothe 7th State of Logistics™ survey releasedin March 2011.

The survey was initiated by the CSIR in2004 to provide decision-makers in publicand private sectors with proper data, infor-mation, etc. about this sector. Secondly, theaim was to quantify important indicators annually to obtain trends over time. Boththese have been achieved and it has beenacknowledged that the State of Logistics™survey, released jointly by the CSIR and its partners in this initiative – IMPERIALLogistics and Stellenbosch University – is an authoritative and important report forthe country.

Topics covered in the 7th survey include‘green’ logistics (including ‘extra distance’measurement); deteriorating road qualityand benefit-cost analysis; supply chain riskmanagement; the skills issue; and rural lo-gistics.

For the seven annual surveys released todate, see www.csir.co.za /sol. – Hilda van Rooyen

| 3 |

OCEANOGRAPHIC INFORMATIONon the southern ocean is crucial tomanaging southern Africa’s longcoastline that is exposed to theweather generated over this vastocean. The highest waves in theworld have been observed in thesouthern ocean, and when the accompanying conditions reachthe coast of South Africa, they impact the beaches and affect safeoperation of ports and shipping.

The Southern African Data Centrefor Oceanography (SADCO) –managed by the CSIR – startedproviding free and open access toits data during 2010. Data aremostly collected from researchvessels, vessels-of-opportunity,moored instruments or floatingbuoys. Much of this information isentered into global numericalmodels, such as those used forweather forecasting, to allow someprediction of climate events and

trends, and even tracking of oilspills.

SADCO forms part of a global network of ocean data centres.These are all part of the United Nations’ IntergovernmentalOceanographic Commission,which was established 50 yearsago. SADCO is funded by a consortium comprising the CSIR,the Department of EnvironmentalAffairs, the South African Navy,the South African EnvironmentalObservation Network of the Na-tional Research Foundation andthe Namibian Ministry of Fisheriesand Marine Resources.

SADCO’s data include more than250 000 vertical profile data setsof the southern ocean, with thecentre being fully computerisedand some data sets dating back as far as 1670!

– Hilda van Rooyen

CSIR helps ‘tame’ southern ocean with research data

infoscopeNews snippets from around the CSIR

MORE THAN THREE-QUARTERSof the world’s cargo is transportedby sea. Eight of the 13 southernAfrican commercial ports are lo-cated in South Africa and ownedand managed by the Transnet National Ports Authority (TNPA).The demand for these ports tohandle greater amounts of ton-nage increases every year.

The CSIR supplies the TNPA with48-hour wave forecasts for itsports and harbours (seven portsfrom Saldanha Bay to RichardsBay). The real-time service pro-vided by the CSIR to Transnet in-cludes the monitoring of waves,tides, currents and wind. This isachieved through the WaveNet

monitoring system(http:wavenet.csir.co.za). The data lead to improvements in the efficient functioningand safety of South Africanports, which handle over90% of South African international trade.

WaveNet obtains data from –among other components –‘waverider’ buoys, which provideenvironmental data crucial forforecast purposes. An access linkwas thus also established for theSouth African Weather Service(SAWS), with data available forthe daily operations of SAWS.

– Hilda van Rooyen

Waves monitored for safer ports operations

A ‘waverider’ buoy,which supplies data tothe WaveNet monitor-ing system

| 4 |

Researchers in Africa to tap into international research on SET data

calculations in all fields of scienceand technology, including physics,biology, geology, astronomy, engi-neering, environmental scienceand ecology.

The expansion of research leads toa corresponding expansion in theamount of data generated. Onethus has to ensure the survival ofphysical as well as electronic datathrough archives. “For such re-search data to be utilised fully,mechanisms must be in place toevaluate, preserve, retrieve anddisseminate such data. CODATAhas a track record of 40 years inthis arena; it is an independentscientific committee of the Interna-tional Council on Science,” com-ments Cooper. CODATA also bringsexperts who do research on dataquality, access, management, etc.,together for collaboration and toavoid duplication of efforts.

Research on aspects of scientificand technology data use is under-taken by task and working groupsin CODATA. One of these commit-tees namely, the newly-estab-lished Data Citation Standards andPractices Task Group, will be de-

“EASY ACCESS to reliable, qualityresearch data is paramount in allfields of science, engineering andtechnology. Increased access to international research data net-works and activities allows for thecreation of new knowledge, dis-semination of existing expertise as well as direct cooperation withpeers abroad,” says AntonyCooper of the CSIR.

Cooper was elected an ExecutiveCommittee member of CODATA,

the interdisciplinary and interna-tional Committee on Data for Science and Technology. As theonly Executive Committee mem-ber from Africa, Cooper says: “It ishighly relevant for researchers inAfrica to increase their utilisationof these international researchdata networks – ease of access is a major consideration.”

CODATA deals with all types ofdata resulting from experimentalmeasurements, observations and

The CSIR’s Dr Martie van Deventer and Antony Cooper who both work towards getting researchers in Africa to tap into international research on science and technology data

| 5 |

Geoportal offers large spatialdata sets at the click of a button

veloping an acceptable standardway of citing the use of scientificdata sets – irrespective of the scientific discipline. The CSIR’s Dr Martie van Deventer is repre-senting the African continent onthe task group. The task groupmembership was a deliberatecombination of expertise from anumber of countries and disci-plines to provide an internationalforum to identify and help recon-cile the needs of various stake-holder communities that arealready working on the onlinedata citation and attribution issues. “Such groups include researchers, academics, informa-tion scientists and publishers,” she says.

“Currently, no international cita-tion standard for online data spansall the disciplines of science andtechnology or could claim generalacceptance. Although we are obvi-ously looking at the online data citation practices already in use incontinental Europe, the UK andthe USA. This is a very opportunetime for African stakeholders toensure that their needs are alsorepresented,” she concludes.

– Hilda van Rooyen

IN LINE WITH international andnational strategies to provide openand free access to earth observa-tion data, the CSIR is now able toserve researchers and externalusers with access to over nineterrabytes of geospatial and earthobservation data.

Operational since 1 April 2011, thenewly-established Geo-SpatialData Infrastructure (GSDI) allowsresearchers from all over the worldto access large geospatial data setsthrough the CSIR’s Geoportal, withthe click of a mouse.

According to CSIR contract research and development man-ager, Laurie Barwell, the CSIR’sGeoportal is in line with interna-tional initiatives such as the Global Earth Observation Systemof Systems (GEOSS) and the SouthAfrican Earth Observation Systemof Systems (SAEOSS) that provideopen access to spatial data and information: “Through the Geoportal, researchers can dis-cover, view, interrogate and mapdata relevant to their research orprojects. This allows our re-searchers to share and interactwith geospatial data locatedwithin a global network of spatial

data providers and users via atruly interoperable system basedon international standards,” he explains.

Enquiries:Geoportal administratorgsdi-ICT@csir.co.zahttp://gsdi.geoportal.csir.co.za

| 6 |

ICT and Society

terms of status, geo-location andability – an expanded Internet towhich services and intelligence areadded. Where the masses of dataare intelligently processed, leadingto enhanced decision-making andactions. The InternationalTelecommunication Union (ITU)has suggested that the ‘Internet ofThings’ and the world’s objectswill be connected in both a ‘sen-sory and intelligent manner’.(International TelecommunicationsUnion, ITU Internet Reports 2005: The Internet of Things. ExecutiveSummary, Geneva: ITU, 2005.)

A new paradigmWhat has made the ‘Internet ofThings’ possible? Thanks to theadvent of cheaper processors andmemory, shrinking sensors andactuators and improved network-ing capabilities, Internet connectedsmart and mobile devices, appli-ances, vehicles and security sys-tems (to name a few) now numberbetween 50 – 100 billion.

“Many of these things are becom-ing smaller and smaller, and inte-grating into everyday life. It is

possible to embed intelligence inthem and connect them to the Internet,” Coetzee notes. “They cansense, compute and act. The infor-mation they provide is integratedwith information from other con-nected things, thus allowing forthe creation of higher order cloudservices available to the connectedthings.” Cloud computing is com-putation, software, data access,and storage services that do notrequire end-user knowledge of thephysical location and configura-tion of the system that delivers theservices.

Another factor contributing to the‘Internet of Things’ is increasingwidespread accessible and afford-able broadband Internet connec-tivity, thus making it easy tointegrate ‘things’ into the Internet.Standards, such as IPv6 (InternetProtocol version 6), and interoper-able communication protocolsbeing created and implemented,aid this endeavour.

Tags and markers provide identifi-cation of non-smart objects thatconnect to the Internet. Throughthis approach, ‘non-smart’ objects

are incorporated into cyberspace.Unique identification by quick response (QR) codes or radio frequency identification (RFID)chips makes it possible to link aphysical object to a cyberspacerepresentation containing informa-tion about such a thing.

Realising benefitsCoetzee sees the real value of thisphenomenon in its potential tosupport decision-making throughcomputing that “melts invisiblyinto the fabric of our business, per-sonal and societal environments”.Through the integration of things,knowledge benefiting society iscreated.

The ‘Internet of Things’ conceptshave been demonstrated in a vari-ety of domains: logistics, transportand asset tracking, smart environ-ments, health, energy, defenceand agriculture. Harnessing the‘Internet of Things’ phenomenonfor power management in a sub-urb is a useful example of its benefits. “By collating patterns ofenergy usage, it becomes possibleto optimise usage, to provide

Connections extraordinary everywhere:‘Internet of Things’

Imagine a world in which our energy demands are optimised, the load managed and controlled through the intelligent control of appliances, utilities and devices. A world where energy consumption is known, where users are empowered to remotely manage their load demand. Imagine a world where the devices at home could make sense of weather predictions and pre-emptively change settings on ‘things’ to make our houses more comfortable?

A WORLD WHERE your alarmclock knows to wake you up ear-lier because of inclement weatherwhich will impact on your transittime to work. Imagine cities withtraffic flowing smoothly, where allthe empty parking spaces areknown and communicated andlinked to your destination – regulated by a vast system of‘things’ that interact with one another to come up with optimalsuggestions for travellers.

Exciting? Definitely, and not thatfar in the future, according to DrLouis Coetzee who leads theCSIR’s research effort on the ‘Inter-net of Things’.

“This is a world where all our devices and physical entities areconnected to the Internet, creatinga network of unprecedented scale,each device or physical entity withits own IP number, each with itsown published set of ‘services’,based on the physical attributesand virtual ‘personality’, allowingus and other devices to interactwith the physical world throughcyberspace,” says Coetzee.These objects are identifiable in

| 7 |

targeted information to that suburb,” Coetzee notes.

Decision-making at a higher levelwill also benefit from the ‘Internetof Things’. “When we combine en-vironmental data on pollutedwater and high rainfall, combinedwith data on deforestation, for ex-ample, it’s fairly simple to extractinformation indicating pollutedflooding, and act accordingly,” Coetzee remarks. He talks of“smart processes and services tosupport economies, environmentand health”.

Research challengesCoetzee admits that while the ex-amples cited above point to thepotential of the ‘Internet ofThings’, there are a myriad re-search questions to be solved.“How do we store the masses ofdata and interpret information outthere? How do we integrate and

interact with so many different devices? Is cloud computing theanswer?”

He believes a fusion of researchtrends – distributed intelligence,statistical analysis and mecha-nisms to ensure interoperabilityand connectedness to heteroge-neous devices and cloud comput-ing – offers the answer. “Ourresearch at the CSIR has demon-strated the success of smaller, ver-tical solutions, such as finding andidentifying misplaced objects orusing QR codes for tree identifica-tion or communicating importantinformation regarding our envi-ronment such as water levels andpotential flooding by using ‘Inter-net of Things’ concepts,” he says.

The concept of smart cities con-jures up interaction between peo-ple and things (man to machine)and between things (machine tomachine). “However, it’s a lot

more complex, as hard informa-tion must be combined with infor-mation on variable and varyingsocial activities,” Coetzee notes.

“For intelligent integrated smartsolutions to become a reality, moreresearch into solutions using dis-tributed intelligence must be done.We need to get to the point wherewe integrate our previous smallvertical solutions.”

Unsolved issues in the context ofthe ‘Internet of Things’ phenome-non are the rights of individuals:Every user of a smart phone is visible through his connectivity tothe Internet! How does this affectpeople? How does one trust a‘thing’? What if incorrect decisionsare made based on informationextracted from data? How doesone plan to avoid this?

Coetzee is keen to tackle thesetechnical and associated social

research challenges with his groupof researchers, as the ‘Internet ofThings’ is a worldwide trend. He isadamant that South Africa oughtto ramp up its capabilities in thisdomain to ensure that local solu-tions can be devised for local problems. “Smart cities utilising‘Internet of Things’ concepts are being planned in differentcountries. Integrated solutions incorporating the physical world,cyberspace and humans are becoming more prevalent.”

He concludes, “It is still early in the‘Internet of Things’ research cycle,therefore we have ample opportu-nity to establish international re-search collaboration and positionSouth Africa in this exciting research domain.” – Biffy van Rooyen

Enquiries:Dr Louis Coetzeelcoetze1@csir.co.za

The term, ‘Internet of Things’,refers to the connectedness of everyday objects (‘things’) to the Internet. Everything will be connected to everything else – any place, any time, any thing, any one, any service, creating unprecedented masses of data to store and process. ‘I

NTE

RN

ET O

F TH

ING

S’

A QR or quick responsecode is a specific matrixbar-code (or two-dimen-sional code), readable bydedicated QR bar-codereaders and cameraphones. The informationencoded can be text, URL(Internet address) or otherdata

“The ‘Internet of Things’brings closer to home thechoices we can make.”

The Internet of Thingsengineering group:Andrew Smith, Sizakele Mathaba,Laurie Butgereit,Promise Mvelase,Guillaume Olivirin, Dr Louis Coetzee andDr Nomusa Dlodlo

| 8 |

Global trendsThe important role of space andICT technology to achieve develop-ment objectives was recognised atthe Millennium Summit (2000)and the World Summit on Sus-tainable Development (2002). Ofsignificance to ICT research, deve-lopment and innovation (RD&I) inearth observation (EO), is the factthat remote sensing and geo-graphic information systems (GIS)were especially highlighted duringthe 2002 Johannesburg confe-rence as important tools in helpinga broad base of stakeholders tomanage natural resources andecosystems. This gave rise to theformation of the intergovernmen-tal Group on Earth Observation(GEO) with South Africa playing aleading role.

Over the last 15 years, earth obser-vation technologies have under-gone tremendous advancement.Coarse and medium resolutionsensors are making way for hyper-

spectral, synthetic aperture radar,ultra-high resolution and hyper-temporal EO platforms developedby major space agencies in response to science and environ-ment-related issues. Constellationsof micro and medium satellites areemerging as promising tools to assess, monitor and provide datain real-time on natural disasters. Since operational reliability hasbeen established in areas of edu-cation, health and environmentalsustainability; it is now widely accepted that remote sensing is animportant pillar of the informationsociety. Investment in the associ-ated information processing, algo-rithm development, innovativeanalytical techniques and man-agement philosophies for largedata sets is necessary to keep pacewith the advances made in thesensing capabilities.

Weather forecasting, climate re-search and land use/ land changedetection and analysis have longdepended on satellite systems for

ICT and Society

ICT research, development and innovation: Absolutely relevant to earth

Over the past 15 years, the convergence between spacetechnology and information and communications techno-logy (ICT) has seen significant maturity. Initially, this convergence was led by satellite-based communications.Over the past eight years, information processing tech-niques from the ICT domain have demonstrated an increas-ing convergence with the processing techniques used onthe terabytes of data collected from the constellation ofsatellites orbiting and monitoring the earth’s surface.

| 9 |

inputs, while benefiting from the ever-improving computing capabilities and modelling and information processing. New research topics borrowing bril-liance from communications signal processing (for example,time series analysis), middlewareconcepts used in transparent networked application develop-ment and web service provision-ing, applied to remote and in situearth observation sensing, are creating paradigm shifts resultingin the next wave of experimentalspace technology demonstrators.

While less focus has been on theoutputs, there has nonethelessbeen an improvement in the ability to produce high quality andreliability digital maps; forecastand predict weather behaviour;characterise dynamic earth sys-tems and patterns; and producefrequent, high fidelity measure-ments of oceans and land change.These advances have resulted inbeneficial social impacts resulting

in space technology transitioningfrom experimental tools into critical infrastructure for nationaldevelopment. Acquiring andmaintaining skills and knowledge,and gaining and maintaining access to information are acceptedas primary enablers to economicopportunity for the poor and marginalised.

The role of GEOSSAgainst this backdrop, one of the GEO initiatives is GEOSS, the Global Earth Observation System of Systems, which is simultaneously addressing nineareas of critical importance to people and society. Its many aimsare to empower the internationalcommunity to protect itself againstnatural and human-induced disasters; understand the environ-mental sources of health hazards;manage energy resources; respond to climate change and its impacts; safeguard water resources; improve weather fore-

casts; manage ecosystems; and promote sustainable agriculture and conserve biodiversity.

The South African National SpaceStrategy places emphasis on mov-ing space technology applicationsfrom experimental to everyday use(operationalisation) to derive societalbenefits, achieve economic goalsand strengthen South Africa’s profile in the community of nations.It is therefore critical that ICT RD&Ibe continued and enhanced amongall the actors within the SouthAfrican space sector, and receivesfurther investment. This will enableSouth Africa and, indeed, other developing countries to succeed intheir efforts towards sustainable social development, mitigatingagainst the devastating effects ofnatural disasters, and reducing thecontributors to climate change.

– Lee Annamalai and Dr Konrad Wessels

observation

“Unlocking information from the masses of satellite data currently available is a major ICT challenge”

Enquiries:Lee Annamalailannamalai@csir.co.za

Dr Konrad Wessels Lee Annamalai

| 10 |

WITH A MEMBERSHIP of 86 national governments and 61 par-ticipating organisations, the Groupon Earth Observations (GEO) hasbecome one of the world’s broad-est and most inclusive forums forEarth observation and environ-mental monitoring.

Better usage of Earth environmen-tal data across disciplinary, geo-graphic and political boundaries isessential for addressing environ-mental crises like tsunamis,droughts, extreme rainfall events

and other environmental impactson humans such as cholera epi-demics. To coordinate the collec-tion and analysis of Earth data,GEO has been leading a world-wide effort since 2005 to build aGlobal Earth Observation Systemof Systems (GEOSS).

As part of these international efforts, the CSIR is involved in several multi-partner internationalprojects to build systems that linkvarious forms of data sets witheach other in a user-friendly and

widely-accessible manner for localand international stakeholders.

One of these, the Earth Observa-tion and Environmental monitor-ing for the mitigation of Healthrisks (EO2Heaven) project, aims todevelop a system whereby differ-ent forms of data sets – from themicrobial to remote sensing – arecoded in such a way as to enableenvironmental researchers to com-bine it with Earth observation datain larger-scale models by the clickof a button (see pages 12-13).

In another project, botanist and re-mote sensing specialist Dr MelanieLück-Vogel is testing the transfer-ability of the methods developedby EBONE in the European con-text for the large scale applicationin non-European Mediterraneanregions like the Sandveld in theWestern Cape. EBONE is the Euro-pean contribution to the GEO Bio-diversity Observation Network(GEO BON), a network dedicatedto the development of standard-ised global biodiversity monitoringsystems.

Lastly, Dr Stewart Bernard is work-ing towards the establishment of aSouth African multi-institutionalmarine remote sensing unit, de-veloping methods according toGEO principles to monitor waterquality and specifically eutrophica-tion in coastal and inland waters(see page 17).

ICT and Society

GROUP ON EARTH OBSERVAT IONS

OTHE

R CO

UNTR

IES OTHER ORGANISATIONS86 Countries 61 Organisations

GEO SECRETAR IAT

>>

> >

SOUTH AFR ICACo-chair of GEO

CEOS

NRF>

SAEON

CS IR SANSA>

SAEOS

> > >

N ine s o c i e t a l b ene f i t a r ea s w i t h s e l e c t ed p ro j e c t s

D i s a s t e r sHea l t hEne rgyC l ima teWa te r

Wea the rE co sy s t emsAg r i c u l t u r e

B i od i ve r s i t y – GEOBON

OTHERSPACE

AGENC I ES>

SANSA

>

SAC

>

NASA

>

ESA

>

From satellites to pathogens –building a global system of systems to understand planet Earth

| 11 |

LEGEND

DST Department of Science and TechnologyNRF National Research FoundationSAEOS South African Earth Observation StrategySANSA South African National Space AgencyCEOS Committee on Earth Observation SatellitesNASA National Aeronautics and Space AgencyESA European Space AgencyGEO BON GEO Biodiversity Observation NetworkEBONE European Biodiversity Observation NetworkEO2Heaven Earth Observation and Environmental modelling

for the mitigation of Health risksEODC Earth Observation Data CentreSBA Societal Benefit AreasUSGEO United States Group on Earth Observation

GLOBAL EARTH OBSERVATION SYSTEM OF SYSTEMSThe plan through which GEO works

GEO PORTAL

>

SAEOSPo r t a l

>

SBAPo r t a l s

>

Other countries’portalse.g.

US GEO

>

EODC

>

Dr Terence van Zyl leads the Sensor Web Enable-ment for In-Situ Observing Network Facilitationtask. The task fosters the development of space-borne, air-borne, sea-based and ground-based sensing networks (sensor webs) and considers scenarios or use cases that demonstrate the valueof sensor webs to the societal benefit areas for example, disasters, health, biodiversity, ecosys-tems and water.

Karel Matthee is editor of the GEONET Mobiletask which supports broader dissemination and distribution networks. It focuses on delivery of information, derived from earth observations data,to people living in rural areas with only access tomobile phone technology.

Dr Stewart Bernard is closely involved with threeGEO tasks focused on developing operational observation systems, as part of the GEO system of systems. The Global Water Quality InformationSystem seeks to develop a fully operational, spatially-comprehensive water quality informationsystem in inland and coastal waters. Contributingnational research initiatives are the CSIR-fundedSafe Water Earth Observation System (SWEOS)project, developing this capability for local pilotecosystems; and aligned DST-funded initiatives tobuild operational capability.

The Chlorophyll Globally Integrated Network(ChloroGIN) aims to broadly disseminate accessiblemarine EO products, and empower users throughtraining. Bernard is closely involved with imple-menting these aims for Africa through EU-fundedprojects such as: the Europe Africa Marine-Network(EAMNet) and GEONETCast for and by DevelopingCountries (DevCoCast).

Finally, the GEO Building Capacity for OperationalOceanography task seeks to develop systems forreal-time ocean observation and forecasting, which is implemented nationally through the multi-institutional OceanSAfrica initiative.

WH Y A G L O B A L S Y S T E M ? ALL OVER THE WORLD, many satellites with sensors to measure changes in theatmosphere, water and on land have been launched; tens of thousands of surface stations have been deployed in the water, on the land, and in the atmosphere usingsensors aboard aircraft to measure and analyse the atmosphere and the land, andwater surface below them on a regular basis. This massive amount of data has the potential to help us in many ways, from weather prediction to disaster manage-ment.

But many of these systems are built independently by governments and organisa-tions for their own purposes, and, in the past, the recurrent problem of a lack of interoperability has made them inaccessible to researchers and decision-makers.

Supported by groups of international teams working to develop technology based on international standards, GEOSS builds upon existing national, regional, and international systems to provide comprehensive, coordinated Earth observationsfrom thousands of instruments worldwide. The resulting system of cooperation provides an international mechanism to broaden the utility of these data into vitalinformation for society in general, as well as domain specific.

In order to become a member of GEO, a government or organisation has to acceptthe basic principles of the sharing of data, information and observation systems.The South African government’s participation has been through the Department ofScience and Technology and it has played a leading role since the establishment ofGEO.Source: Earthzine available at http://www.earthzine.org

| 12 |

THE RAISON D’ÊTRE of thesemodels will be to assist environ-mental managers and decision-makers to identify and thenmitigate the impact of environ-mental health risks such ascholera.

“It is very difficult to understandan ecosystem from only one per-spective. For example, as a micro-biologist, I can identify a cholerabacterium in a water sample. Butthat does not mean there is goingto be a cholera outbreak tomor-row,” explains senior researcher,Dr Martella du Preez.

In order to be able to provide anearly warning system, for exam-ple, researchers need to charac-terise an entire catchment, andonly once they have modelled it,can they start to explore whatwould happen when environmen-tal conditions start changing insuch a way as to be conducive to acholera outbreak. A cholera epi-demic, for example, is driven by avariety of environmental factorssuch as rain, temperature and UVlight, as well as human-relatedfactors such as a settlement’sproximity to water sources, DuPreez explains.

Combining and analysingdifferent data setsFunded by the European Union,the Earth Observation and Environmental modelling for themitigation of Health risks(EO2Heaven) project aims to takeone step further and develop asystem whereby data sets fromenvironmental researchers can becombined with other data setssuch as remote sensing imagesand other Earth observation data.

According to Du Preez, her teamgenerates massive amounts of

data: “In the Loskop Dam andGroblersdal area, for example, wehave over two years of data onwater quality, air and water tem-perature, land-use practices andmore, but the data remain lo-calised to the specific samplingsites and no-one else has access toit.

“If we link this data set with Earthobservation data we can start tounderstand what is happening ona catchment level scale. In thesame way, Earth observation andremote sensing data are often notuseable until someone has geo-

ICT and Society

Microbiologists, data analysts, modellers and computer scientists are joiningforces to harness the advantages offered by new technologies and Earth observationdata to build dynamic models of ecosystems and their environmental characteristics.

| 13 |

located them precisely with physi-cal reference points on the groundin a process we call ‘ground-truthing’. In some instances, weliterally take the samples on theground, at the same moment thesatellite passes over the samplingarea,” she explains.

However, transforming the micro-biological data into a product thatcan talk to other data sets is noeasy task. Data sets from varioussources need to be coded and thentranslated using algorithms so thatyou can compare them with an-other data set, often using ad-vanced statistical techniques,explains modelling expert Marnavan der Merwe.

With a background in microbio-logy, Van der Merwe acts as thelink between the microbiologistsand the computer specialists whoare building the information archi-tecture so that microbiologists like

Du Preez only need to press a but-ton to get access to these data setsand experiment with the vari-ables.

“As soon as you have a data setcovering an entire catchment, re-gion, or a country like Uganda,then we can go really small andproduce maps of those areas mostimpacted by cholera, without los-ing sight of what is happening oncatchment or regional scale,” saysVan der Merwe.

“A satellite cannot identify orsense what kind of pathogensthere are in the water – for that weneed the microbiology data fromthe field. However, when linkingthe microbiology and water qual-ity data sets with earth observa-tion data, we can see howenvironmental factors like tempe-rature, rain and ultra violet radia-tion play a role in the onset of acholera epidemic,” she explains.

Challenge to computer scientistsFor Grame McFerren and DrAnwar Vahed, computer scientists,the most important aspect is thebreaking down of silos. “We haveisolated, fragmented and incoher-ent systems that limit the use ofvery expensive data sets, collectedby public institutions like the CSIR.

“The most difficult part of this pro-ject is not necessarily the scienceor development of the technology,but changing people’s attitudes.Often researchers confuse custodi-anship of data sets with owner-ship. Data sets are not intellectualproperty – it should be madeavailable to the wider community.

“It is also extremely expensive togenerate these data sets. It istherefore important that we coor-dinate efforts to prevent duplica-tion.”

The cholera case study is a first attempt to present environmental-health data in an OGC compliantway to benefit society.

Computer scientist and systemsarchitect, Dr Terence van Zyl con-curs: “The challenge is not techni-cal, but people. Fortunately wehave a wonderful team of re-searchers on this project, eager totry new things and opening up to

From satellites to pathogens – mapping the environmental drivers ofcholera epidemics on a larger scale

the possibilities a system like thiscan offer. “In this regard, we areteaching the Europeans a fewthings!” he laughs.

Van Zyl explains that even thoughenvironmental researchers andmicrobiologists will not need toknow too much about informationand communications technology(ICT) to operate the system, theystill retain control over their datasets and how it is used in modelsand maps by means of the scien-tific workflow concept (see page15).

“Previously, environmental researchers would carefully sam-ple an area for two to three years before they then compile the datato produce a static map that is truefor those three years. For any fu-ture work they would have to gothrough the whole process again.“With our system, once a catch-ment or ecosystem has been modelled, you only need to feed it with the latest field data to seehow it is changing. You can alsogo back and change the para-meters and see how that’s influ-encing the system. Once you’vemodelled a catchment, it is somuch easier for us as humans tounderstand what is happening tothat system,” Van Zyl explains.

–Wiida BassonEnquiries:Dr Martella du Preezmdupreez@csir.co.za

WHAT ARE OPEN GEOSPATIAL CONSORTIUM (OGC) STANDARDS?

The EO2Heaven system is based on internationally-accepted OpenGeospatial Consortium (OGC) standards. OGC is the ‘standards’ body forthe Earth observation community and their standards have been adoptedby the Group on Earth Observations (GEO) to standardise data sets inter-nationally – a concept termed ‘data democracy’. In other words, if datasets and systems conform to OGC standards, you will be able to link it tothe Global Earth Observation System of Systems (GEOSS).

The South African response in support of GEOSS is embodied in the SouthAfrican Environmental Observation Strategy (SAEOS). The South AfricanEnvironmental Observation Network (SAEON) portal will provide an entrypoint to the SAEOS platform.

| 14 |

What is AMESD?It is a continental-wide, pan-African project for the develop-ment of geo-information services.Its aim is to improve decision-making processes in the fields ofenvironmental resource and envi-ronmental risk management inAfrica. This is done by increasingthe information management capacity of African regional andnational institutions mandated forenvironment-related sectors, andby facilitating access to Africa-wide environmental informationderived from earth observationtechnologies.

Leading the CSIR’s participation inAMESD is Philip Frost, a CSIREarth observation specialist whowas responsible for developing theAdvanced Fire Information System(AFIS) together with Eskom. Histask now is to use an innovativecombination of information andcommunications technologies(ICTs) to ensure that fire servicesand products developed at theCSIR are replicated in the SADC.Frost explains, “The fire service ispart of the thematic action(THEMA) for the SADC region,with the focus on agricultural and

environmental resource manage-ment.”

An agricultural service anddrought service are also plannedin addition to the fire service.”These two services are being ad-dressed by the Agricultural Re-search Council in South Africa.

The nature of the services to beprovided in line with the SADCTHEMA was determined througha series of consultative SADCworkshops with stakeholders.Needs expressed by national ministries of agriculture and of the environment have had to bematched by services.

Novel ICT combinationsThe novelty of the AMESD projectrelies in part on the manner bywhich information is dissemi-nated. Frost points out, “Dissemi-nation is completely Internetindependent via a communica-tions satellite. The GEONetCastdissemination portal requires theCSIR to provide daily products viaa file transfer protocol server toGermany. These products are thenpushed up to the communicationssatellite and downloaded to receiv-ing stations.

Frost is overseeing the roll out of20 receiving stations, one for eachministry of the environment in theSADC region, “The AFIS terminalwill be used as part of the receiv-ing stations.” Through remote access, software will be continu-ously downloaded to ensure thatreceiving stations remain in primeworking condition. Fixed meshantennas of 2.1 m are the secondvital component of these receivingstations to be installed.”

Novel ICT productsFrost ensures that the SADC receiving stations have access tothe CSIR’s live fire alerts and firedanger alerts. “These products aredeveloped from the data receivedby the CSIR Satellite ApplicationsCentre and are constantly beingvalidated,” he confirms. Active firealerts are displayed as the mainscreen; these are derived from themoderate resolution imaging spec-troradiometer (MODIS) and MSG.The fire danger product is derivedfrom the fire weather index (tem-perature, humidity, rainfall andwind) by using the conformal-cubic atmospheric (CCAM) model.The CCAM is a variable-resolutionmodel, developed by CSIRO mod-ellers in Australia, and is run oper-

ICT and Society

Enhancing African environmentalmanagement through ICTThe CSIR’s expertise in developing fire products has seen it take on an important role in the South

African Development Community (SADC) economic region as part of the African Monitoring of theEnvironment for Sustainable Development (AMESD) project. AMESD is the outcome of an

investment by the European Development Fund (EDF) and a partnership between the African Union (AU) Commission, the European Union (EU) Commission, the five

African Regional Economic Communities and the Secretariat of the African, Caribbean and Pacific Group of States (ACP).

Promoting informed decision-making for environmental monitoring in Africa

| 15 |

ationally on the computers of theCSIR Climate Studies, Modellingand Environmental Health group.The operational weather forecastsare produced early-morning everyday at a horizontal resolution of 15 km for the next 7 days. Thethird product, the burnt area estimate, is supplied monthly toassist in calculating areas affectedby fire; this can in turn be used to calculate carbon dioxide emis-sions.

Helping to realise AMESDthrough trainingThe CSIR has taken on the joint responsibility for training repre-sentatives of SADC ministries inthe use of the fire service, alongwith the universities of Botswanaand Zimbabwe who provide thetraining for agriculture anddrought services respectively. This involves immersing traineesin theory and case studies, as well as the writing of manuals.

– Biffy van Rooyen

THE ‘TRIED AND TRUSTED’method of documenting scientificdata and methods in laboratorymanuals is slowly but surelychanging as scientific workflowsfor various domains gain traction.This according to Dr Terence vanZyl, who is part of the team thatleads the CSIR’s research on scientific workflows for earth observation (EO).

What are scientific workflows?A scientific workflow system is aspecialised form of a workflowmanagement system designedspecifically to compose and exe-cute a series of computational ordata manipulation steps, or aworkflow, in a scientific applica-tion. What makes scientific work-flows so valuable is the fact that aprocess followed during scientificwork can be reused and repeated,and has provenance. “The samepiece of software can be used for adifferent application,” Van Zylpoints out. "In scientific work-flows, provenance refer to captur-ing of the steps and alternativepaths followed, as well as the results obtained during data ex-ploration."Having this informationat one’s disposal provides a his-tory of the success and/or failurethat constituted the scientific endeavour,” he notes. “This meansthat one can revisit the work andtake alternative paths from thework to explore different aspects.”

Scientific workflows forearth observationVan Zyl’s team is keen to provethat scientific workflows can beused operationally for EO science.“Scientific workflows can bridgethe gap between EO and statisticaltoolsets, such as R, which havespecific applications,” he explains.

“We are also working to narrowthe gap between geospatialtoolsets, such as GRASS andPySAL, and scientific workflows.”R, GRASS (Geographic ResourcesAnalysis Support System) andPySAL (a library of spatial analyticfunctions based on the Pythonprogramming language) are allopen source toolsets used in theEO domain.

The CSIR’s research focuses on information and communicationstechnology (ICT) for EO and operates within the workspace of the OGC (Open Geospatial Consortium) and its Sensor WebEnablement (SWE) framework.The OGC is an international voluntary standards organisationencouraging development and implementation of open standardsfor geospatial content and services. SWE standards enabledevelopers to make all types ofsensors, transducers and sensordata repositories discoverable, accessible and usable via the web.“We are planning to bring toolsfrom this workspace into scientificworkflows,” Van Zyl confirms.

Van Zyl and his team have deve-loped a package in VisTrails, a scientific workflow managementsystem developed at the ScientificComputing and Imaging Institute

at the University of Utah in theUnited States. VisTrails providessupport for data exploration andvisualisation. EO4VisTrails is apilot for their research in prepara-tion for making it operational. The team aims to have a spatialanalytical demonstrator up andrunning soon.

In conclusion, Van Zyl emphasisesthe value of scientific workflows insupporting innovation. “When wedevelop a scientific workflow todetermine, for example, how closehouses are to the coast, it becomespossible to reuse this science effec-tively for many applications, suchas decision-making regardingrisks and vulnerability.” Risks andvulnerability are relevant to busi-ness, environmental managementand government, to name a few.

The scientific workflows team regards the larger scientific community who would derivevalue from this research as their‘clients’. Van Zyl is looking for-ward to future collaboration on scientific projects locally and inter-nationally. – Biffy van Rooyen

Enquiries:Philip Frostpfrost@csir.co.za

Decision-making for environmentalmanagement: EDF investment in Africa

2002 – 2005 Project for the Utilisation of Meteosat Second Generation(MSG) Satellite in Africa(PUMA) €11 m

2010 – 2013 AMESD €21 m

2013 – 2016 Monitoring of the Environmentfor Security (MESA)

“It’s like a high-tech

lab notebook”

Scientific workflows in support of Earth observation science

Enquiries:Dr Terence van Zyltvzyl@csir.co.za

For more information, visithttp://code.google.com/p/eo4vistrails/

The EO4Vistrails team: (from left, clockwise): Dr Terence van Zyl, Bheki Cwele. Bolelang Sibolla, Dr Anwar Vahed, Derek Hohls, GraemeMcFerren and Petrus Shabangu

| 16 |

THE SOLUTION IS BASED ongraphical processing unit (GPU)hardware obtained from multina-tional specialist supplier NVIDIA ata fraction of the cost of central pro-cessing unit (CPU) high perform-ance computing.

Used in more than 90% of newdesktop and notebook computers,a GPU is a specialised micro-processor that offloads and acce-lerates image computation fromthe CPU. Modern GPUs are very efficient at manipulating imagesand their highly parallel structuremakes them more effective thangeneral-purpose CPUs for a rangeof complex algorithms.

“The architecture of the compactNVIDIA Tesla s2050 system iscalled SIMD (Single InstructionMultiple Data) and contains 1 796processing cores for parallel dataprocessing and is highly suited toprocessing complete images usingthe same algorithm,” according toRiaan van den Dool, GroupLeader: Advanced Image Process-ing at the CSIR.

“We are experiencing a significantincrease in processing speed be-tween in-house developed algo-rithms running on CPU and thoseusing GPU hardware. The remark-able difference in the time taken

by the two processes is illustratedin the figure above, which reflectsa speed-up of roughly 700 times.This performance bodes well forprocessing large amounts of datausing complicated image process-ing algorithms, which will enablethe CSIR to timeously deliver pro-ducts dependent on order-drivenprocessing,” says Van den Dool.

The newly-acquired GPU hard-ware complements the data andinformation management systemor DIMS, which integrates theCSIR’s SARMES1 processing suiteand several other processors. Thiscreates an advanced customersearch and order interface.

“The growing number of receivedsatellites, diversity of data and demand for products continuouslydrive the need to stay at the cut-ting-edge of highly automateddata and processing management.Our multi-mission EO data management system approachgives us the flexibility to do this.We are excited about the new GPU capability and the improvedspeed with which we can processimagery for our customers,” saysWolfgang Lück, a technical manager at the CSIR.

ICT and Society

High performance computing revolutionises earth observation image processingA recently-acquired high performance computing solution has revolutionised the processing speed of remote-sensing products fromsatellite over-pass data with as much as 700 times.

The running time comparisonbetween in-house developedalgorithms running on theCPU (left) and GPU (right)hardware as shown heredemonstrates the remarkableincrease in processing speed.

Riaan vanden Doolnext to theNVIDIA system andits hostserver (anHP DL380)

1 South African Resource Management and Expert System developed in-house at the CSIR by a technical manager, Wolfgang Lück.

Enquiries:Riaan van den DoolRvdDool@sansa.org.za

The Earth Observation (EO) unit at the CSIR Satellite Applications Centre centralised the CSIR'sEO services into a spatial portal that acquires,archives, processes and distributes remotely-senseddata from EO satellites. The data are processed asremote-sensing products and imagery and providevaluable input for a broad range of decision-makers.

GPUs at their bestThe utilisation of GPUs in HPC is gaining momentum. Three systems in the TOP10 ofthe TOP500 list of the world’s fastest super-computers are based on GPU architecture,which confirms how fast GPU systems can be. The Number 1 system on the TOP500,Tianhe-1 (self-built by the National DefenseUniversity of Technology in China), is basedon the GPU architecture.

The Centre for High Performance Computing(CHPC in Cape Town) has a GPU cluster of 20NVIDIA cards as a test-bed for users to testtheir applications. NVIDIA is an Americanglobal technology company which specialisesin the development of graphics processingunits. Good acceleration of the in-house CFD (computational fluid dynamics) CODE developed by the CSIR’s Dr Arnaud Malan has been achieved. An increase in applica-tions benefitting from accelerators such asGPU may result in a national system builtaround this architecture.

| 17 |

WHEN CSIR RESEARCHERSneeded new types of equipmentwith which to gather scientificdata from water bodies to studythe occurrence of harmful algaeand eutrophic states, they called in assistance from their researchengineering colleagues in thefields of mechatronics and sensorscience.

Currently, there is little under-standing of the mechanisms bywhich eutrophic states occur, andunder-developed national infra-structure to measure eutrophica-tion in real-time. To better observeand fully understand eutrophica-tion scientifically, data are needed,preferably in the form of real ornear real-time systems coupledwith new data products fromEarth observation satellites. Butthe equipment and methods withwhich to gather data are prohibi-tively expensive and not alwayssuited to South African conditions.

CSIR oceanographer, Dr StewartBernard, explains: “The Benguelacurrent off the south-west coast ofSouth Africa is regarded as one ofthe most productive ecosystems inthe world. The inland waters wework in also can contain very highconcentrations of algae, particu-larly noxious cyanobacteria. Ourlocal conditions and systems aretherefore very demanding forsatellite and buoy-based measure-ments, and require the develop-ment of new techniques,platforms and sensors.”

For South Africa to have this kindof technology at hand and at lowcost, will have a huge impact onour ability to monitor water qualityissues such as red tide on the west coast or problems withcyanobacteria and eutrophicationin inland waters: “Currently, marine research vessels cost about

R100 000 a day to operate, andare unable to provide daily obser-vations, while commercially-avail-able autonomous measurementsystems can be prohibitively expensive or unsuitable for localconditions. Earth observationdata, coupled with measurementsfrom critically-located buoys, canoffer very low-cost data in nearreal-time across many differentecosystems,” Bernard says.

The CSIR’s Safe Water Earth Observation Systems (SWEOS)project’s aim is to build Earth observation-based informationsystems to provide products forecosystem monitoring and charac-terisation, ultimately facilitating apredictive capability. The newbuoys will be used both to providereal-time data on algal blooms,and to develop and verify newsatellite techniques, which willthen allow satellite data recordsfrom 10 years or longer to be usedto characterise change in SouthAfrican coastal and inland ecosys-tems.

The buoy solution, according tothe CSIR’s Johan Olivier, who project-managed and undertookthe mechanical design work onthe system “is a pencil buoy specifically developed for waterobservations. It needed to have a low-cost, low-weight designwhich is easy and fast to deploy”.In addition to the buoy, the CSIR’sOptronics group is developing newlow-cost radiometric sensors, withthe aim of greatly reducing systemcosts.

The CSIR’s Peter Bosscha, respon-sible for the electronic and soft-ware development work, explains:“The buoy’s electronics have avery low power design which canstay deployed for a long time with-out requiring maintenance. Solar

panels are used to keep the buoy’sbattery topped up and data are col-lected at regular intervals. Datacan be stored on board but canalso be relayed via a GSM (cellphone) network to a centralserver.”

The buoy can be used for bothocean and inland water monitor-ing, and initial trials were held atHartbeespoort Dam in September2010 with a prototype assembly.After further mechanical, elec-tronic and software work, thebuoy is currently undergoing seatrials in the Southern Benguela,one of the world’s most productiveupwelling systems. During thesetrials, algal blooms are monitoredwhich can have a detrimental ef-fect on fisheries. The next stop willbe Loskop Dam in Mpumalanga.

The buoy, affectionately dubbed‘Gizmo’, is equipped with sensorsfor radiometry, fluorescence/backscatter, temperature, wind speed,water current speed and GPS.

The development of the new buoy and sensor also serves as‘proof of concept’ that the CSIRhas the capability to design andbuild low-cost robotic observationsystems for aquatic ecosystems.The ultimate aim is to developnew technology that allows hundreds or even thousands of robotic sensors to provide much-needed environmental data for inland, coastal and deep oceanecosystems.

New systems developed to study water quality and eutrophication

Eutrophic states can occur inboth inland and ocean waterbodies when elevated nutrientlevels occur. It can lead toalgal growth which may betoxic or greatly reduce the dissolved oxygen levels inwater. The result is the deathof plants and creatures thatare natural and vital to theecosystem.

Enquiries:Dr Steward Bernardsbernard@csir.co.za

The CSIR-developed buoy dubbed ‘Gizmo’ was tested at Elands Bay on the south-west coast from 23 February to 14 March 2011 by a team of scientists from the CSIR, the Department of Agriculture, Forestry and Fisheries and the University of Cape Town.

The CSIR-deve-loped SWEOSbuoy bobs in atest pool at theCSIR before beingshipped to thecoast for tests inthe ocean. Onlyhalf of the buoyis visible abovethe water.

| 18 |

“AS THE ENGINES of nationalgrowth, South Africa’s cities needto create jobs and provide ade-quate shelter, infrastructure andservices to alleviate poverty andimprove the health and safety pro-file of their inhabitants while stim-ulating vibrant economies andpromoting more sustainable, en-ergy-efficient urban settlements,”says Dr Louis Waldeck who headsthe CSIR’s Urban Dynamics Labo-ratory.

The populations of these cities aregrowing rapidly and like in manyother countries, many SouthAfricans are moving from ruralareas to cities in search of a betterlife. During 2005, 59.3% of theSouth African population lived inurban areas compared to theglobal average of 48.7%. It is ex-pected that 71.3% of the SouthAfrican population will be ur-banised by 2030.

South Africa’s four largest metro-politan areas and their immediatesurrounding regions are home tomore than 38% of the nationalpopulation. The growing and di-verse economies of these metro-

politan areas generate more than66% of the national economy andform the backbone of SouthAfrica’s economy.

“The CSIR laboratory provides theexperimental space within whichlikely patterns of urban growthcan be studied 30 years into thefuture, thus allowing planners anddecision-makers to evaluate thelong-term impact of spatial poli-cies and investment decisions thatour large cities are making today,”comments Waldeck. The labora-tory brings together a trans-disci-plinary team and a numericalmodelling and simulation plat-form that look at cities, regionsand their development trajectoriesas complex, adaptive social-eco-logical systems.

The current simulation platform isbased on various open-sourcesoftware projects. UrbanSim isused for micro-simulation of thechoices made by individual house-holds and businesses as con-sumers of property and services,developers as suppliers of prop-erty, as well as government as asupplier of infrastructure and ser-

vices. MATSim simulates the be-haviour of transportation systemsfrom individuals pursuing theirdaily activity schedules. “Currentindividual choice capabilities in-clude route choice, activity timing,activity and activity locationchoice, mode choice and transferbetween modes (e.g. park andride), toll avoidance and informa-tion feedback response (e.g. radiotraffic reports),” notes Waldeck.A major obstacle to applyingmicro-simulation and agent-basedtechniques in the developingworld is the availability of data asthese techniques require informa-tion about individual households,jobs, buildings, vehicles, etc. Someof this information can be derivedfrom the 10% sample of Census2001 forms made available to theCSIR team by Statistics SouthAfrica (Stats SA).

“This allows us to generate a ‘syn-thetic’ population that is statisti-cally representative of the realpopulation in a city. We have anexcellent working relationshipwith Stats SA, who has gone outof its way to make special extractsavailable to improve the quality of

our ‘synthetic’ populations. For re-searchers in our field, Census 2011is very exciting, as we will have anew baseline for our simulations.Other micro-level information canbe deduced from the zoning andvaluation-roll information kept bymunicipalities,” explains Waldeck.As part of its drive to harness spa-tial information and communica-tions technology (ICT) in support ofdevelopment agencies to bringabout improved quality of life, theDepartment of Science and Tech-nology (DST) commissioned theintegrated development planningand modelling project (IPDM). Thedevelopment of the IPDM, aimedat supporting integrated planning,development and service deliveryfor South Africa, is undertaken bythe CSIR and the Human SciencesResearch Council, in collaborationwith various universities.

The IPDM not only involves thefirst major application of theCSIR’s simulation capability in thecities of eThekwini, Johannesburg,Nelson Mandela Bay and the en-tire Gauteng Global City Region,but an adaptation of technologiesoriginating from a developed

ICT and Society

Research aids sustainable planning of

30 years aheadThrough urban simulation expertise, CSIRresearchers assist metros with enhancedplanning in and around cities, includingservice delivery and infrastructure develop-ment, for up to 30 years into the future.

| 19 |

world context to uniquely SouthAfrican circumstances. Further de-tails about the IPDM project areavailable at http://stepsa.org.The CSIR’s Maria Coetzee, projectleader of the IPDM, explains thatthe results of the simulationprocess will enable metros andcity regions to plan more proac-tively for the long term by betterunderstanding:

• Future patterns of demand forinfrastructure, facilities andservices such as water, electri-city, sanitation, schools, clinicsand hospitals.

• The investments that will be required to sustain the economy, including public and private transport.

• How future urban forms mayimpact the sustainability of ourcities by using indicators suchas travel time and costs, accessto social and economic oppor-tunities, and energy and car-bon efficiency.

The IPDM team uses a living labo-ratory approach, involving theplanners and city decision-makersof South Africa’s major metropoli-tan areas in a series of interactivework sessions to develop, test andapply the ICT platform in real-lifecontexts of respective cities or cityregions.

“Some of the main activities of theliving laboratories include the de-velopment of various demo-graphic, employment and spatialplanning policy scenarios to besimulated by the UrbanSim/MAT-Sim platform. This will enableplanners in these metros to assessthe possible future impacts of theircurrent growth managementstrategies and spatial frame-works,” concludes Coetzee.

With the Tshwane city landscape in the background, Dr Louis Waldeck, manager of planningsupport systems at the CSIR and research group leader Maria Coetzee

Simulated residential urban expansion at land parcel scale between 2001 and 2007

Enquiries:Dr Louis Waldecklwaldeck@csir.co.zaMaria Coetzeemjcoetze@csir.co.za

| 20 |

UNTIL RECENTLY, BROWSING,navigating, filtering and process-ing of information on the inter-web could only be done throughtext-based language. However,Google has pioneered a techno-logy that allows people to accessthe web using the spoken word.Through its new technology,Google has launched a servicedubbed ‘Voice Search’ that en-ables Internet users to do Googlesearches by speaking into their cell phones, making it possible tosearch on Google using voice inthe same way as one wouldsearch by typing in a search query. Users download an application ontheir smart phone and call the In-ternet to obtain information. Onaverage, one voice search is about100Kb and uses an acoustic, pro-nunciation and language model toproduce accurate results based onsearch queries.

Enter the CSIR and North-West UniversityVoice Search is available in SouthAfrica in South African English,isiZulu and Afrikaans because of

the work done by the CSIR’s andNorth-West University’s humanlanguage technologists.

Human language technologistsrange from engineers, computerscientists, statisticians, mathe-maticians to linguists, sociolin-guists and sociologists. They applytheir expertise in the multidiscipli-nary field known as Human Lan-guage Technology (HLT).The CSIR has its own team of re-searchers who specialise inhuman language technologies.“The HLT group’s focus is reallyabout applying science to enablepeople to interact with machinesusing natural language,” says DrKaren Calteaux, operations man-ager at the CSIR’s HLT group. “Weare particularly motivated by re-search that is relevant and applica-ble to South African languagesand ICT for development,” sheadds.

HLT researchCalteaux indicates that the team ofHLT researchers investigate howHLT can be adapted and applied to

benefit users in developing coun-tries. According to Calteaux, HLTresearch rests on three major pil-lars namely: developing speechtechnologies within a multilingualenvironment and for resource-scarce languages; seeking real-world situations where thesespeech technologies can be tested,applied and promoted; and grow-ing future human language tech-nologists. The group continuallyproduces top-achieving students,says Calteaux. “Five of the Mas-ter’s degrees emanating from thegroup between 2009 and 2010,were awarded cum laude,” sheadds.

She defines HLT as an enablingtechnology that can play a crucialrole in addressing social cohesive-ness and development by empow-ering people with access toinformation. “Telephone-basedsystems using HLTs such as auto-matic speech recognition and text-to-speech synthesis now allowpeople to access information inmore than one language usingvoice, which is ubiquitous,” sheadds.

HLT in action in SAIn South Africa, HLT is used in efforts to support language diver-sity in an affordable and equitablemanner, explains Calteaux. HLT is also used to assist industry andgovernment to make services anddocuments available in all 11 offi-cial languages. Thus HLT has amajor role to play in rectifying thehistorical discrimination againstspecific languages.

In this regard, the HLT team hasbeen involved in Project Lwazi, aninitiative co-funded by the depart-ments of Arts and Culture (DAC),and Science and Technology(DST). The aim of Project Lwazi isto increase the impact of speechtechnologies in South Africa byshowing how these can enhanceand facilitate socioeconomic development through telephone-based, speech-driven informationsystems. – Bandile Sikwane

ICT and Society

CSIR and North-West University language technologists together with Google,

pioneer the spoken web in South African English, isiZulu and Afrikaans

Enquiries: Dr Karen Calteauxkcalteaux@csir.co.za

HLT operations manager, Dr Karen Calteaux

|21 |

WHAT MADE IT POSSIBLE forWatson to compete against andbeat its human competitors byproviding quick responses to rapidquestions with hidden meanings,puns and riddles? Part of the ex-planation lies in the use of ontolo-gies. This field of research at theCSIR is led by Dr Tommie Meyer.

Ontologies and knowledge repre-sentation enable the developmentof semantic technologies (seman-tic pertains to meaning in lan-guage), by which meaning isencoded to make it possible forpeople and machines to under-stand, share and reason using

knowledge. This differs greatlyfrom hard-coded information usedin traditional information techno-logy. For the latter, queries can retrieve only facts that have beencaptured and stored, but do notmake connections or inferences(deriving logical conclusions) tocome up with new information.

Using the logical inference that ispart of formal ontologies, it be-comes possible to come up withnew information from the facts thathave been captured. Herein lies thedifferentiator of ontologies: It offersa way to derive value from existinginformation in a particular domain.

Ontologies have already proved tobe of value in the domain of gene-tics as a component of a unique research method being used tosolve complex problems.

So, which domain is up next for the ontological treatment? Two new domains to benefit from ontology research are biodiversityand ecosystems, both of which arecited as part of the 2010 NationalGovernment Performance Measure-ment Outcomes: ‘Environmentalassets and natural resources thatare well protected and continuallyenhanced’.

How to create an ontology for a domain

Dr Aurona Gerber, who is passion-ate about ontologies, explains theprocess of creating an ontology.“First we need a shared, consen-sual and conceptual domainmodel, as it’s all about talking the same language as the domainexperts,” she says. Agreed termi-nology and shared vocabulary areessential.

Domain experts in natural sciences define and share expertterminology. The role of the onto-

Uncovering new knowledgein scientific domains

using ontologies

In February 2011, Watson, the IBM supercomputer, took the

crown as Jeopardy winner. Watson had beaten two of the most

successful players ever to have taken part in this American

quiz show featuring trivia in history, literature, the arts, pop

culture, science, sports, geography, wordplay and more. The CSIR's Aurona Gerber, Tommie Meyer and KodylanMoodley are excited by the potential that ontology tech-nologies promise, enabling people and machines to understand, share and reason using knowledge

Logical inferencing:

New facts can be inferred from asserted facts through logical consequence, for example: • In summer, all days are hot.• On a hot day, it always rains.• Therefore, it rains on all

summer days.

| 22 |

ICT and Societylogy researcher is to capture this real knowledge in a formatthat is accessible to a computer through the use of logic-founded knowledge representation.

Once this process has been completed, the knowledge representation is checked for consistency. “Now comes thereally exciting part,” Gerber notes. “Our conceptual modelhas thousands of concepts and ten times as many relations.When we add data to the concepts, new classifications become possible, thereby adding new knowledge to the domain.” The equivalent outcome through fieldwork, experimentation and other methods such as research algorithms, often takes years to achieve.

Gerber is excited when confronted with the issue of massesof data stored in different databases, “It’s theoretically possi-ble to couple these repositories with domain models to inte-grate information at the semantic level.” Advanced queryingmechanisms can then be investigated to exploit complex relations.”

Integrating results from these research efforts can be used todevelop application solutions to provide decision support forpolicymakers.

Biodiversity and ecosystems: Important, innovative, timely and relevant research

Gerber is confident that biodiversity and ecosystems infor-matics is a future fruitful area of research. She explains why this application of ontologies holds so much promise,“Semantic models with ontologies that capture and reasonabout meaning can assist scientists to understand the inter-relationships of the domains. This has already been done inmedical informatics with ontologies such as SNOMED CT1.Knowledge integration between relevant domains at thelevel of meaning would be very valuable.”

Gerber and her colleagues are keen to apply the same scientific rigour to the domains of biodiversity and ecosys-tems to assist some of South Africa’s key researchers at theCSIR and other national institutions. “At present, there islimited use of semantic technology in these domains. Intelligent information systems can add greatly to current efforts to manage these precious resources,” she remarks.

Potential impact is envisaged as protection of South Africa’sbiodiversity heritage to facilitate sustainable development,thereby protecting ecosystems services on which SouthAfricans depend for their livelihood and life services. Ontolo-gies can also facilitate the potential for bioprospecting andassist in the protection of intellectual property.

This will, in turn, boost South Africa’s standing as a countrywith leading research and experts in biodiversity andecosystems. – Biffy van Rooyen

Enquiries: Dr Aurona Gerberagerber@csir.co.zahttp://www.ihtsdo.org/snomed-ct/

Empoweringcommunities tobuild networks

‘Empowering communities to build

their own networks’... This is the philosophy that

holds firm in the BB4All (Broadband for All) initiative,

a joint initiative between the CSIR, the departments of Science and Technology (DST), and Communications.

| 23 |

TO THRIVE IN TODAY’S WORLD,it is important to understand therole ICT plays in society and thedevelopment of people. BB4Allaims to provide affordable broad-band connectivity to under-ser-viced, rural communities. It useslow-cost infrastructure that isowned and supported by the localcommunity, to create socio-eco-nomic and commercial opportuni-ties. At the heart of the BB4All initia-tive is the Wireless Mesh Network(WMN) project. It is funded by theEuropean Union through the Sec-tor Budget Support (SBS) pro-gramme of DST. The collaborative project started in2009, with roll out of a broadbandnetwork in the Nkangala Region.The project will in coming monthsroll out to the Sekhukhune Districtin the Limpopo Province and theJohn Taolo Gaetsewe District in theNorthern Cape.

The local heroes

The key success factor in this ini-tiative is the Village Operators(VOs), who work within the com-munities to provide and managethe BB4All network. There arecurrently 19 VOs working in 15communities in the Nkangala re-gion. VOs are young local entre-preneurs with a keen interest inICTs and who want to be self-em-ployed. They are sourced from thetargeted communities and will

service their own communities. To date, wireless mesh equipmenthas been installed in more than200 facilities and is currently beingconnected to the Internet. These fa-cilities include 175 schools, schoolcircuit manager offices and a com-munity radio station.

The initiative uses established wire-less mesh research and transformsit into an innovative approach toreal-life situations. Through this intervention, the improvement ofthe quality of life of rural or under-serviced areas of South Africa hasbecome a possibility.

This project also provides the oppor-tunity for multidisciplinary researchwhere social and business aspectsare integrated with technological research to provide a sustainablebusiness model.

The VOs are encouraged throughtraining and mentoring to build and maintain their own businesses.They receive business and basictechnical training, which is aug-mented by training on WMN tech-nology and experiential learning.All these aspects support and encourage self-learning, a key skill for the VOs and communitiesto be self-sustainable.

The BB4All model learns fromother past and current CSIR ICT innovations and models such as Infopreneurs® and the Digital Doorway (see page 44).

Broadened skills base inwireless mesh technology A local manufacturer has beencontracted to manufacture some ofthe wireless mesh equipment. Par-sec manufactures the high per-formance nodes, a key componentin WMN technology. This stimu-lates the use of locally-manufac-tured products and local labour.Ingwapele Technologies, headedby Louis Bapela, employs 10 per-manent staff members, who aretrained to install the nodes andother equipment unique to WMN.This is contributing to broadeningthe local skills base of the wirelessmesh technology and to skillsbuilding of these young people.

Sustainability for now and into the futureThrough training and develop-ment, the BB4All initiative aims toleave a legacy in this community,a business model powerfulenough for the community to sus-tain and develop further. The VOswill be well-equipped with re-sources, knowledge and the con-tacts to grow their ownbusinesses, support their commu-nity and be the real advocates ofICT.

BB4All is a growing initiative withambitious goals. At its completion,it will have connected at least 450

facilities (schools, clinics and othergovernment facilities) and deve-loped at least 45 VO businesses.Broadband infrastructure in ruralareas will help with access to es-sential services such as health careand education, and bring otherservices, such as governmentservices, closer to people. Theseare communities that were previ-ously almost marginalised due todistance and circumstance. As ICTcitizens of the world, they havethe potential to become partici-pants in the global information so-ciety.

Nombulelo Molumaela, one of theVOs attests, “For me, the BB4Allinitiative brought a lot of know-ledge. I didn’t know about broad-band and wireless technology; I am now informed and will inturn share this knowledge withmy community. In this way I willbe adding value to my communityjust as the initiative has addedvalue to my life.”

Another VO, Sibusiso Mazibukoadds, “We (the VOs) wake upevery morning to go to the office.You should see Michael Mabenaand Innocent Nene going to theiroffice; you would be convincedthey are working on Wall Street!Such is the pride we have in thisproject.” – Bonang Tselane

Some of the VillageOperators from theKwaMhlanga area,spent time with theCSIR's ICT experts toget to grips with theirnew responsibilities

Enquiries:Kobus Rouxkroux@csir.co.za

BB4All management team (front row from left): Ajay Makan, Alida Veldsman, Kobus Roux, Neil Oosthuizen Backrow: Marlien Herselman, Ronel Smith, Madelein van den Berg, Bonang Tselane, Mario Marais

| 24 |

THE INFOPRENEUR® concept is arural innovation and developmentmodel that guides and supportsthe establishment of primarilyyouth-owned social enterprisesthat form a sustainable, informa-tion network. Given its successfultrack record, the project is nowlooking forward to collaboratingwith institutions, individuals orcorporations that share its values.These include the ‘Ubuntu’ ap-proach to business, social entre-preneurship, developmentthrough enterprise, virtual andphysical networking, effective in-quiry, and ICT-promoting cultureand community development.

The information networks estab-lished by the Infopreneur® con-cept, help rural communities toovercome the digital divide andchallenges related to being ‘off thegrid’. These challenges include in-sufficient access to governmentservice delivery; ineffective accessto extensive trade opportunitiesinto and out of communities; theabsence of access to modern tech-nology, connectivity; and the infor-mation that such connectivitywould provide. An Infopreneur® isan information entrepreneur.

There are currently 20 Infopre-neurs® distributed among theVhembe, Bojanala and Siyanda(Namakwaland) districts. The Info-

preneur® project has evolved overthe years, from identifying, train-ing and skilling individual Infopre-neurs® to assisting them inestablishing social enterprises. In-fopreneurs® are in the business ofcollecting information in theircommunities, to be used by thecommunity and for external par-ties.

Apart from servicing communitieswith information, the success ofthe network is also based on theprovision of services to govern-ment (provision of data on eco-nomic activities), the private sector(a channel to market directly to aninformal or previously untappedmarket) and developing commu-nities (e.g. SME business services)in a development-through-enter-prise approach. Research chal-lenges have been in thetechnology adoption and owner-ship spheres as well as the twinissues of scalability and sustain-ability.

As community-based ICT-en-abled, sustainable, micro serviceenterprises, Infopreneurs® act asthe steering mechanism thatdrives community development,by giving community membersaccess to information gatheredfrom different sources, and help-ing them work with this informa-tion.

ICT and Society

Information services to support entrepreneurial development

Infopreneurs® assist with a varietyof information. Two current exam-ples are a bee farming project andanother is the National RecordalSystem (see next article).

Hive of informationInfopreneurs® is helping youngentrepreneurs establish their ownbusinesses in the bee farming in-dustry. The young businessmenand women will be farminghealthy bees from the KrugerPark, harvesting the honey andpacking it for sale. Infopreneurs®assist entrepreneurs directly or putthem in contact with specialistswho will train them in bee keep-ing, hive management and honeyproduction and marketing. An example of using ICT innovativelyin this context is the practice ofadding Quick Response (QR) codeson honey labels. These QR codesare readable by a smartphone,and when photographed, directthe enquirer to information provided on the Internet by theentrepreneurs. Information includes the location of the hive,the number of bees in the hive,and the location and origin of thehoney. – Bonang Tselane

Enquiries:Johann (Rensie) van Rensburgjvrensbu@csir.co.za

Infopreneur with a rural entrepreneur

What is indigenous knowledge?

“The local knowledge that istacit, expressive and unique to a given culture or society and is the basis of local-level decision-making including, but not limited to agriculture,health care, food preparation,education, natural resourcemanagement, economics, governance and security.”

| 25 |

AN EXCITING and challenging initiative to record, preserve, protect and promote SouthAfrica’s invaluable wealth of IK is being researched, developedand implemented by the CSIR onbehalf of and in close collaborationwith the Department of Scienceand Technology’s (DST’s) NationalIndigenous Knowledge SystemsOffice (NIKSO).

Implementing the national IKS policyThe National Recordal System(NRS) initiative, spearheaded byDr Yonah Seleti, Chief Director ofNIKSO, supports communities,guilds and other IK holders inrecording their IK with the ulti-mate aim to create opportunitiesfor benefits to flow back to thecommunities. These benefits mayinclude both social and economicbenefits.

The NRS implementation encom-passes the establishment of IKNetworks, IK documentation centres and an ICT knowledgeplatform, all of which will bestrengthened by appropriateprocesses and stakeholder involvement to achieve the aimsencapsulated in the IKS policy.

In addition, the initiative aims tobring selected fragmented data-bases together through a portal toenhance the National System ofInnovation for research and deve-lopment purposes within the national priorities of the DST.

Innovative and new technologieswill be introduced to facilitate theNRS’s aim to collect un-capturedIK (traditionally transferred fromone generation to another in anoral and tacit format) in multime-dia without losing too much of thecontextual information. New and/or enhanced existing processesand systems will respect the cultural, traditional customs and etiquette of IK holders andcommunities while capturing,managing, storing, protecting andresponsibly publishing IK informa-tion.

NIKMAS: the digital brain of the NRSIn support of the NRS initiative,the National Indigenous Know-ledge Management System(NIKMAS) will be a semantic digital repository to be used as the technology platform that will,among others, manage prior artdeclaration.

NIKMAS is being developed inparallel with the IK collection activity in the communities, inorder to support the new processesand requirements identified anddeveloped by and within the communities. Although the deve-lopment of NIKMAS poses severalsignificant technical challenges, ithas the potential to be the leadingIK management system both nationally and internationally.

Safeguarding the future ofindigenous knowledge through ICT Cultural riches are transferred from one generation to another through indigenous knowledge (IK).Unlike India where IK is documented, for example, South Africa’s IK vests largely in oral traditionand is running the risk of becoming extinct as a result of westernisation and urbanisation.

Dr Yonah Seleti (DST NIKSO), Carol van Wyk (DST NIKSO), Riëtte Pretorius, Hennie Bezuidenhout (both CSIR)

| 26 |

NIKMAS is far more than just acatalogue of IK resources, the CSIRNRS project leader HennieBezuidenhout explains, “NIKMASwill be an intelligent knowledgerepository.” It comprises a seman-tic digital repository with custom-developed meta data schemataand a sophisticated security modelto protect and preserve the IK, anadvanced semantic search engine,a sophisticated cataloguing systemand an overarching integration architecture that combines thesubsystems into a coherent, fit-for-purpose system.

Intelligent reasoning through ontology research on IK-relatedmeta data will be encapsulated inthe design and development ofNIKMAS. Ronell Alberts, a CSIRresearcher studying towards herMSc at the Free University ofBolzen-Bolzano in Italy will complement this ambitious andnovel system by focusing her studies on the use of ontologiesand semantic web technologies forintelligent searching, specificallyon indigenous knowledge.

The first version of NIKMAS willfocus on the African TraditionalMedicine and Indigenous Foodsthemes as dictated by the nationalpriorities of the DST. Other themeswill be prioritised and imple-mented together with the appro-priate stakeholders over the nexttwo years. It is envisaged thatNIKMAS will provide significantsocial and scientific benefits tocommunities and researchersalike.

IK networks and centresThe recordal of IK is a complexprocess requiring a two-prongedapproach: The cataloguing of IKholders (who, what and where)and the actual recording of know-ledge.

According to Carol van Wyk, DST Deputy Director: KnowledgeManagement, much of SouthAfrican IK is still in oral format andby establishing IKS Documenta-tion Centres (IKSDCs), the depart-

ment’s aim is to effectively enableand to facilitate the collection of IKwithin communities by communi-ties, to be captured in NIKMAS forthe primary benefit of communityand research opportunity withinthe National System of Innovation.The IKSDCs are viewed as pillarsfor the NRS in local and indige-nous communities. Two comple-mentary models (i.e. communitynetworks and IKSDCs) are cur-rently being tested in the commu-nities to define an appropriatemodel for the recording of indige-nous knowledge.

IK is being recorded through pilotnetworks in the provinces ofLimpopo, North West and theEastern Cape. Through these pilotnetworks, recording systems andprocesses are tested and refined ina Living Lab context. In addition,community members are sourcedand skilled to act as IK recorders tocollect information from their com-munity members.

At present there is one pilot IKSDCat the University of Zululand(UniZulu). This pilot is being usedto test and scope the ideal IKSDC.This and future pilots will be implemented and tested byNIKSO.

Giving back to communitiesAn action-based research method-ology was adopted to allow forconsistent and continuous learn-ing from and for the communities.CSIR researcher Riette Pretoriuspoints out, “Communities areguided and supported to structurethemselves. They are skilled andtrained to develop business plans,and on how to use various ICT andother electronic equipment, for example, cameras and global

positioning system devices.”Pretorius continues, “Above all,through guidance by NIKSO, theyare informed of their rights andtaught that their rights are protected through various legaldocuments that form part of the IK recording process such as memorandums of agreement,prior informed consent, informa-tion transfer agreements and non-disclosure agreements.”

Bezuidenhout confirms, “Our approach ensures practical measures for IK holders and practitioners to develop serviceswith a particular focus on IKS with a scientific base.”

Responsible exploitationAs part of the envisaged valuechain and to ensure that socialand economic benefits accrue tothe IK owners (communities andIK holders), all captured IK datawill be securely stored in NIKMASwith appropriate security controlmechanisms to manage access toinformation. Pretorius reveals, “Intellectual property policies, procedures and processes deve-loped by the relevant authoritieswill drive the protection and exploitation of IK, while NIKMASwill secure the IK knowledge captured.”

Ultimately, the NRS will provide a structured process by which cul-ture can be promoted, sustainablelivelihoods supported and know-ledge shared. – Biffy van Rooyen

Enquiries:Hennie Bezuidenhouthbezuide@csir.co.zaRiëtte Pretoriusheaston@csir.co.za

ICT and Society“NRS is envisioned as the leadingindigenous knowledge systemstreasure hub for South Africa’scommunal socio-economic development.”

| 27 |

SOUTH AFRICA’S investment ine-infrastructure acknowledges theincreasingly digital nature of theresearch and has put in place asystematic way of addressing thechallenges of e-Science, throughlarge investments such as theSouth African National ResearchNetwork and the Centre for HighPerformance Computing (CHPC).Building on these, a federation hasbeen formed by most of SouthAfrica’s top universities, nationallaboratories and research groups,to provide a framework for colla-boration. The South African National Grid (SAGrid) provides acoherent platform for the integra-tion of computing resources, butalso human networks, such as technical experts and scientific collaborations within SouthAfrica’s borders and well beyond,encouraging the full utilisation ofthis common investment.

The investments made so far bythe Department of Science andTechnology (DST) – SANReN (seepage 30), the CHPC and the upcoming initiative to curate largedata sets – has unlocked a certainamount of potential in scientific

and industrial advancement,bringing to certain parts of theresearch community tools whichthey would not previously havehad. SANReN has certainly provena game-changer in this respect,bringing about a potential seachange in the way that peoplework and communicate. With in-stitutes on the network deploying,integrating and jointly managingcompute resources, the full poten-tial of the network is unlocked.

Dr Bruce Becker is responsible forthe SAGrid which is managedfrom the CSIR. He explains howthe SAGrid encourages full use of the benefits of South Africa’scyberinfrastructure to the benefitsof all sub-groups and individuals.“This has had huge benefits toseveral research communities, interms of efficiency, collaboration,cost-effectiveness, competitive-ness, and scale,” Becker says. “It isthe means by which geographyand institutional capacity are nolonger factors in perpetuating inequalities, but become accessi-ble to all researchers who may usethem to their own competitive advantages.”

The fundamental aspect of thegrid is that it provides a coherentinterface to a vast array of servicesfor users. These include the tradi-tional HPC services as well as thewell-known high-throughput andhigh-capacity computing, but extend to data curation; data management and movement;metadata catalogues; identity; and security services. High-levelservices such as workflow engines, compute agents, softwaremanagement, monitoring, accounting and collaboration services are also a fundamentalpart of the grid. Another definingfeature is that no matter where theparticular resource is or what itsphysical details, the user and application can always interactwith the service in a standard way.

“This also means that we canshare our infrastructure and ser-vices with other similar initiativesin the world, opening up vast opportunities for collaboration. Recently, such integration hasbeen done with the sites in NorthAfrica, with a Memorandum ofUnderstanding (MoU) with theEUMed-Support project, and a

ICT and Human Capital Development (HCD)

Making the most of infrastructure:The role of the SAGrid

Today’s research is overwhelmingly digital –whether it is related to data or computing or anycombination of such services in a collaboration.

similar MoU will soon be signedwith the European Grid Infrastruc-ture as well as the GISELA projectin Latin America. These infrastruc-ture sharing agreements of coursemean that all South African re-searchers have access to not onlyincredible compute power, but itgets better: the human networksbuilt by these technical collabora-tions mean that user and applica-tion support, training andapplication porting are part of thepackage. So, no matter whereusers are or with whom they areworking, we can use a commonplatform to support them and theirapplications. The benefits of thisare manifest and will soon be feltby several institutes in Africawhere new sites and applicationswill come on board,” Becker states.

SAGrid provides all of the servicesnecessary to conduct science forthe 21st century. It is open to all resource providers and users. Visit www.sagrid.ac.za to find out more. – Biffy van Rooyen

Enquiries:Dr Bruce Beckerbbecker@csir.co.za

SAGrid enthusiasts: Dr Bruce Becker, Albert van Eck (Deputy Director: IT, FreeState University) and Christiaan Kuun(SANReN project manager at the CSIR)

The Brain Gain InitiativeThe Brain Gain Initiative (BGI)started with a pilot project in2008, and is now in its secondyear of activity. Managed by UNESCO, the project aims to provide a bulwark to the cripplingbrain drain experienced by Africa.This is done by creating researchand collaboration opportunities toAfrican diaspora, through thepower of grid computing.

The CSIR is involved in the project through the South AfricanNational Grid (SAGrid) which provides technical support and coordination to this project. Themembers of this project representsome of the best research groupsfrom the continent and the selected research projects range from air quality to e-waste management. All of them have astrong need for e-Infrastructure.

This is where SAGrid (see page 27)comes in. In collaboration with theEGEE (Enabling Grids for E-sci-encE) project (now evolved intoEGI (European Grid Initiative)),SAGrid has provided training,technical support and operationalsupport to the project itself, as wellas to the individual participants. The CSIR’s Dr Bruce Becker who

coordinates the SAGrid, com-ments, “With the start of theCHAIN (Co-ordination and Harmonisation of Advanced e-Infrastructures) project and thedevelopment of the AfricaGrid Regional Operations Centre, weare starting to see many more sitescoming onto the grid.

“While SAGrid is coordinated fromthe CSIR and one of the sites onthe grid is actually hosted at theCSIR (ZA-CSIR-C4), it is the na-tional operations team – consist-ing of the IT experts at the varioussites who form the federation –which is the real power behind the grid.

As the technical partner to theBGI, Becker was tasked to providetraining to a user group (scientistswho would be using the gridnodes for research) and a technicalgroup (who will be required to operate the grid nodes). Joininghim on the SAGrid team of trainerswere Albert van Eck, Deputy Di-rector: IT at the University of FreeState, and Dr Valeria Ardizzone, an expert in grid computing fromthe INFN (Instituto Nazionale diFisica Nucleare) in Italy, and tech-nical coordinator of the DECIDE(Diagnostic Enhancement of Con-fidence by an International Distri-

ICT and HCD

| 28 |

An African e-infrastructure The CSIR places a very high premium on collaboration and human capital development at all levels of research, from the high school all the way throughto the laboratory. This is evidenced by its involvement in two projects whichhave recently made headlines, both sponsored by Hewlett-Packard (HP) Corporation’s Global Social Innovation Office.

The time is now ripe for an African

e-Infrastructure that would help provide African

solutions to Africanproblems for the

overall developmentof the continent.

Deputy Director: IT from Free State University, Albert van Eck (seated) andthe CSIR’s Dr BruceBecker

| 29 |

buted Environment) project, whohas been working with Beckersince the inception of SAGrid in2008.

“During the training session, theGridOps team has proved its worthby not only running some of thetraining sessions for the 38 partici-pants, but also ensuring that thevarious services were in perfectorder, and coordinating the deploy-ment of new sites across the conti-nent.

“Thanks to the knowledge transferenabled by SAGrid, some of the ex-perts at these new sites will in turnprovide manpower to the regionaloperations team, thereby increasingthe capacity of the grid and ensur-ing uptake and sustainability of theinfrastructure.”

Marc Bellon and Liliana Simionescu ofUNESCO were responsible for theBGI project planning of the Pretoriaevent. This follows on from an ear-lier African pilot project. Belloncommented, “The willingness tolearn among the group of Africanresearchers was matched by the acquisition of new skills. Similarly,we are pleased that the under-standing of the technical group hasbeen enhanced through new anduseful information.”

HP Catalyst InitiativeThe second project that the CSIR isinvolved in is the Catalyst Initia-tive, a global network of consortiathat brings together leading edu-cational institutions and NGOs, toexplore new approaches to sci-ence, technology, engineering andmath (STEM) education.

In 2010, the CSIR Meraka Institutewas awarded a grant as the leaderof the Global Collaboratory consor-tium, one of the HP Catalyst Initia-tive’s innovation themes. TheGlobal Collaboratory unites the facilities, minds and skills of CairoUniversity, Faculty of Computersand Information (Giza, Egypt);Coventry University (Coventry,UK); Del Mar College (CorpusChristi, Texas, USA); East CarolinaUniversity (Greenville, North Carolina, USA); Masinde MuliroUniversity of Science and Techno-logy (Kakamega, Kenya) andStamford Public Schools (Stam-ford, Connecticut, USA).

This collaboratory enables stu-dents to participate in collaborativeproblem-solving to address urgentsocial challenges using the powerof collaborative grid computing.Becker points out, “The collabora-

tory links almost all of the humancomponents of a healthy researchenvironment, all the way from thelearner, through the teacher, to the researcher. It brings to the environment state-of-the-art infrastructure and resources,which can be incorporated in theresearch and learning experience.By ensuring that these are rele-vant in the learning experience,we are preparing the students fora life in research, which is a verystrong determining factor in acountry’s economy.”

Several other South African insti-tutions are active in the HP Cata-lyst Initiative, notably NelsonMandela Metropolitan University,University of KwaZulu-Natal andNorth-West University.

New opportunities and challengesBecker sees new and interestingchallenges ahead for the HP Cata-lyst and BGI, “While working atthe CSIR, we put the finishingtouches on the sites in Senegaland Morocco which are now on-line and in the process of being in-tegrated. In the coming months,we will be doing training and con-ducting scientific workshops in

in the making

The Brain Gain Initiative will deliver a durable, valuable asset – an infrastructure and a community of users.several key countries. I will behosted by the Ecole Nationale desSciences de l’Informatique in Tuniswhere we will run a scientificworkshop for distributed comput-ing, together with UTIC (Unité derecherche en Technologies de l’In-formation et de la Communica-tion).

“The SAGrid core services teamfrom the University of CapeTown’s Information and Commu-nications Services department willbe going to Kakamega and Nairobiin May 2011 on invitation to finishthe integration and run dissemina-tion workshops with the vice-chancellors of their respectiveuniversities.” – Biffy van Rooyen

Adele Botha is part of the New Learnerconsortium (led by theAgastya FoundationinIndia) of the CatalystInitiative. She is work-ing on a peer-learningsystem – one of theoutcomes of the project

Enquiries: Dr Bruce Beckerbbecker@csir.co.za

Team members of the Brain Gain Initiative

| 30 |

THE SOUTH AFRICAN NationalResearch Network (SANReN) hasachieved significant milestones inits implementation of a researchnetwork for South Africa. Its latestachievement has been the com-pletion of the Durban metropolitannetwork and the Pretoria DWDM(dense wavelength division multi-plexing) ring.

The intervention is rapidly reach-ing its chief objective which is toconnect research and higher edu-cation institutions (HEIs) in Preto-ria, Johannesburg, Bloemfontein,Cape Town, Port Elizabeth, EastLondon, Durban and elsewhere inthe country on a 10Gbps (Gigabitsper second) optic fibre ring net-work.

This network will enable the re-search community to engage inmeaningful online collaboration. Itwill be used to link to internationalbandwidth acquired for researchpurposes.

A sea change in connectivityWhy is SANReN set to change theresearch landscape in South Africaradically? Prof Colin Wright wholeads the SANReN team, explains,“SANReN increases local band-width for its users and providesthem with international band-width at an affordable rate. Itmakes for better spending of fi-nancial resources.

“By providing research networkingto South African research andHEIs, we help these institutionsfree up funds that might other-wise have been spent on purchas-ing bandwidth for researchers.”

TENET was responsible for negoti-ating a good price for 10 Gbps ca-pacity on SEACOM, a fibre opticcable providing high capacitybandwidth to southern Africa,East Africa, Europe and South

Asia. This is distributed throughSANReN which integrates theavailability of the local and inter-national bandwidth. Geographi-cally diverse routes are currentlybeing investigated by TENET forredundancy of bandwidth.

Another benefit of SANReN – withprofound indirect future impact –is the uptake by a generation ofnew researchers of the benefits of-fered by improved bandwidth forresearch traffic. Bandwidth abun-dance resulting from SANReN’snetworking of universities has thepotential to shape the growth anddevelopment of a new generationof students whose knowledge andskills will contribute to the goal ofcreating an inclusive informationsociety. This in turn will enable socioeconomic benefits throughinformation and communicationstechnology – specifically broad-band.

In turn, these advances on the scientific front will contribute tothe competitiveness of local indus-try through the scientific break-throughs achieved and throughthe establishment of the world-class national cyberinfrastructure.

Bandwidth ups the antefor astronomy and spacescience To date, SANReN has helped toensure that South Africa takes itsrightful place among its peers inthe radio astronomy in terms ofVLBI data provision to the Joint Institute for VLBI in Europe. Dataon radiation from space are gath-ered at HartRAO, the only radio astronomy facility in South Africa,and are sent to Dwingeloo in theNetherlands. “South Africa istherefore able to take its place inthe international radio astronomyexperiments,” confirms Wright.

Unlocking the potential of research networking

ICT and HCD

2003Conceptualisation in partner-ship with the then Department of Arts, Culture, Science and

Technology (DACST)

2006Research infrastructure

becomes DST budget item

Contract between DST and the CSIR

2007National implementation

phase commences

2008National research

cyber-infrastructure live in GautengTI

MEL

INE

The SANReN team of the CSIR: (Front) Prof Colin Wright, Simeon Miteff, John Hay, Thabo Koeshe and Christiaan Kuun (Second row) ZukisaniMakalima, Uli Horn and Geoff Daniell

| 31 |

The aim of these experiments is touse correlation of VLBI data to es-tablish a picture of the planets.

SANReN connectivity will ensurethat South Africa has the requisitebandwidth to tackle the SquareKilometre Array (SKA) telescopeproject. The SKA is set to revolu-tionise radio astronomy. This newgeneration astronomical facilitypromises a larger radio wave col-lecting area than any other facilityin the world; it will be 50 timesmore sensitive and able to surveythe sky 10 000 faster than anyother imaging radio telescopearray.

South Africa’s own microsatellite,SumbandilaSAT, also benefitsfrom SANReN’s capabilities. Dataare downloaded by the CSIR Satel-lite Applications Centre and sent toSunspace in Stellenbosch. Receiv-ing information in a short space oftime enables engineers to makethe necessary adjustments to the

microsatellite for its optimal per-formance.

Value additionsSANReN has successfully con-cluded a South Africa eduroampilot to allow local researchers andstaff from participating institutionsto access the Internet while visit-ing another participating institu-tion via wireless hot-spots.

‘eduroam’ is a worldwide roamingservice. Through the use of RA-DIUS proxies, eduroam allows log-ging on using the credentials fromhome institutions, regardless ofthe institution being visited. SAN-ReN began the pilot in 2010 to es-tablish the proxies required forlocal institutions to participate.Several institutions participated inthe pilot.

In February 2011, the SANReNeduroam RADIUS proxies wereconnected to the European

eduroam root. This temporaryconnection allows for global test-ing of eduroam in South Africa, inanticipation of the establishmentof an African eduroam root. Visi-tors to South Africa can useeduroam at all eduroam-enabledinstitutions. Additional futureservices for user participation arebeing developed.

What’s next?The SANReN team will continue toroll out the network to reach allhigher education and researchsites throughout the country. Atthe moment focus is applied onextending the network to ruralsites, ensuring that ‘no one getsleft behind’. The SANReN teamwill seek to secure a national dark-fibre network, as such a network will provide for almostlimitless national bandwidth, as well as long-term bandwidthsecurity for the research and education community.

It will also become involved in assisting with the adoption of newnetwork technologies in SouthAfrica by using its fibre networks totest equipment capable of 40-100Gbps speeds per wavelength. Thiswill be of benefit to the communitysince the learning obtained can beused to build the next generation ofthe SANReN network.

In addition, the team will focus onproviding more services through thenetwork similar to eduroam. It willensure that the network can beused to provide maximum benefitto its users by supporting servicessuch as data recovery sites, cloudservices, video conferencing plat-forms and e-learning, to name justa few. – Biffy van Rooyen

FAS T FACTSSANReN is part of the national cyberinfra-structure initiative. It complements theCentre for High Performance Computingand the Very Large Datasets curation project

Function: To provide networking and network services to research and higher educationinstitutions

Funder: Department of Science and Technology

Implementer: CSIR, as a strategic initiative

Partner: Tertiary Education Network (TENET) operates the SANReN network

2009National research network

tender to Telkom

National backbone network completed

2010Planned connection to

SALT and SKA

2011Pretoria DWDM ring network goes live

CSIR SANReN steering committee members Laurens Cloete and Johan le Roux

Enquiries:Prof Colin Wrightcjwright@csir.co.za

| 32 |

ICT and HCD

THE INTER-GOVERNMENTALagreement signed by the Depart-ment of Science and Technology(DST) and the Finnish Ministry ofForeign Affairs in July 2007 isdrawing to a close. The SouthAfrica-Finland Knowledge Part-nership on ICT (SAFIPA) kicked offits programme in June 2008 withthe objective of supporting the creation of an environment conducive to the development and deployment of ICT service applications for South Africans.

The programme, which is man-aged by the CSIR on behalf of theDST, had three strategic goals:human capital development; technology research and innova-tion enhancement.

SAFIPA is now in its final year.Kristiina Lähde, chief technical advisor, and Thiru Swettenham,SAFIPA national programme coordinator, say that in terms of itsmain three goals the programmehas been ‘quite successful’.

With great success comesgreat responsibilitiesSays Lähde: “Our success hasbrought about some new chal-lenges in that our stakeholderswant to know who will continuewith the good work that SAFIPAhas done.” She adds that as theprogramme nears completion, oneof the main priorities that emergedis to find a local organisation thatwill adopt and replicate some ofSAFIPA’s projects.

“Some of the projects are simplerthan others, for example we’ve de-veloped a training programme forICT and mobile entrepreneurs;these have been well received. Toadopt and replicate such a pro-gramme is easier than finding an

organisation to provide a similargrant fund to SAFIPA’s.”

For operational and grant fundingpurposes, the SAFIPA programmereceived R3 million from theFinnish Ministry of Foreign Affairs,and R9 million from SouthAfrica’s DST. In order to solve thechallenges that have arisen due toSAFIPA’s success, the steeringcommittee and supervisory boardand the two governments haveagreed to extend the programmeto December 2011.

The three components ofSAFIPA’s programmeTo achieve the overall objective of the SAFIPA programme, threeinterdependent components weredesigned.

Institutional capacity-developmentCapacity building involved valueadded instruction, the training oftrainers, activities with multipliereffects, and networking. It also involved institutional and humancapacity-building.

Innovative IT applications and new solutions for end users The projects identified addressededucation, learning sector and societal development issues suchas empowering disabled andaging populations. All of these had specific human capital development outputs among the target constituency.

The sustainability of the solutiondelivered was considered very important and hence special attention was attached to the identification of the institutionaleducation and learning outputs.This component had two ele-ments: selection, upgrading andimplementation of the projects

already identified; and identificationand preparation of new projects andstakeholder groups.

Network creation and disseminationComponent 3 aimed to strengthencooperation between research institutions both locally and globally, as well as to support public-private partnerships in theservice delivery process.

For dissemination purposes, theprogramme organised or supportedan annual international conferencetogether with relevant programmestakeholders such as universities,companies and communities to inform programme results, inno-vate new development activities,and promote networking.

Recognition where it is dueSAFIPA-supported projects havesince caught the eye of the NationalScience and Technology Forum(NSTF). As a result, SAFIPA-sup-ported projects have been nomi-nated in two categories of the NSTFawards: ‘Research leading to Inno-vation by Teams or Individualsthrough Organisations’ and ’Communication for Outreach and Creating Awareness of Science, Engineering, Technologyor Innovation’.– Bandile Sikwane

SAFIPA’s curtain drawing

Enquiries:Thiru Swettenhamtswettenham@csir.co.za

Kristiina Lähdeand ThiruSwettenham

| 33 |

HPC LIES IN THE SPECIALISTdomain where cutting-edgecomputing equipment and as-sociated software components– predominantly open source innature – come together to pushthe boundaries of what can beachieved in terms of processing,networking, storage and visual-isation.

“Supercomputers help us solveproblems today instead of to-morrow,” states Sithole. “Theirability to perform computationsvery rapidly and handle ex-tremely large volumes of dataeffectively makes them invalu-able for research projects wherelarge volumes of data must beprocessed.”

This particular class of prob-lems requires semi-infinite

computing resources. The possibi-lity to simulate and model results is the most important advantagethat supercomputers have. This inturn supports industrial competi-tiveness, which depends to a largemeasure on planning for the future today – and well-estab-lished industrial competitiveness is a precondition for economicgrowth.

Using HPC to achieve national benefits

Through its flagship programme,the CHPC has enabled researchersand industry participants toachieve results in selected fields ofapplication. Although these fieldsof study and application rangefrom astronomy and aeronauticsto electronic engineering applica-

ICT and the Economy

FAST FACTS ABOUT THE CHPC

• Launched in 2007• Provides users with access to Sun Microsystems; Blue Gene/P & iQudu• Houses the fastest supercomputer in Africa• Has an extensive human capital develop-ment programme in collaboration withAIMS, South African universities and European & African partners

High performance computing (HPC) has been described

as the Ferrari of the computing world – elegant, power-

ful and very, very fast. Thanks to the investment by the

Department of Science and Technology (DST) in the na-

tional Centre for High Performance Computing (CHPC)

as part of the national infrastructure, these benefits are

now available to South African industry, academia and

science councils. The CHPC is based in Cape Town, with

director Dr Happy Sithole at the helm.

Reaping competitive advantage through high performance computing

and data curation

Dr Happy Sithole of the CHPC

ICT and the Economy

| 34 |

tions and oceanography, severalstand out for the novelty of theHPC application and the potentialfor HPC to achieve results in thesepriority areas.

Energy storage materialsThe increasing scarcity of fossilfuels and their soaring cost havenecessitated and accelerated thesearch for alternative energysources. A flagship project led byProf Phuti Ngoepe of the Univer-sity of Limpopo used computa-tional modelling techniques on theCHPC’s IBM cluster to complementexperimental approaches in thestudy of high-power lithium ionrechargeable batteries for cars andrenewable energy resources.

The work focused on the simula-tion of the complex microstructureof lithium manganese oxide (lowcost, environmentally-inoffensiveand with high energy density material) for use in high capacitycomposite electrodes. These simu-lations have laid the foundationfor modelling current and futurebattery compounds of interest,particularly those microstructuresused to predict vehicle perform-ance.

Ngoepe’s group engaged with Optimal Energy, responsible forthe manufacture of the first elec-tronic vehicle in South Africa. Thegroup has assisted with batterytesting and battery choice for theprototypes that preceded Joule,South Africa’s first electric car.

Computational biomechanicsIn a research effort to understandthe mechanics of the heart, agroup of biomechanical scientistsat the University of Cape Town isalso using the power of HPC. Led by Dr Thomas Franz, re-searchers and students are model-

ling the behaviour of the mechan-ics of the heart on the CHPC facili-ties – with and without theaddition of a plastic and chemicalcompound used to repair weak-ened heart muscles.

This synthetic polyethylene glycolhydrogel is being investigated byDr Neil Davies of the Cardiovascu-lar Research Unit to build up heartwalls following a heart attack. After it has been injected, the hy-drogel forms a web-like substancethroughout the damaged tissue.By comparing the behaviour of the heart mechanics, it becomes possible to isolate other factorsthat may be playing a role in theheart repair process.

This is the second CHPC flagshipin life sciences, a trend that islikely to continue as studies in lifesciences have the potential to impact positively on the popula-tion at large.

Creating art with HPCSouth Africa’s first full-length ani-mated movie by Character Matters,‘The Lion of Judah’, hit the screenin September 2010. The facilities ofthe CHPC were used to produce it.

South Africa has now joined othercountries, such as New Zealand,where HPC facilities are used formovie-making. The CHPC wasable to assist Character Matterswith the rendering of the movie,which requires working with hugevolumes – 25 terabytes – of data.These data are needed for the virtual production of animatedmovies.

“We hope to assist aspiring SouthAfrican filmmakers by providingaccess to our facilities,” Sitholestates.

He confirms that apart from theseapplications in energy, medicineand the animation industry, other

opportunities abound for theCHPC to support the competitive-ness of other industries, such asthe local petrochemical industry,as well as the financial sector.

Data curationThe Very Large Database (VLDB)initiative, the third leg of the na-tional cyberinfrastructure invest-ment by the DST, envisages thestorage and curation of researchdata from all research domains,particularly in areas of environ-mental and climate change model-ling, bioinformatics, medicalsciences and astronomy. “Provid-ing large data repository and cura-torship platforms enablesustainable knowledge sharingand other benefits for cyberinfra-structure,” Sithole points out.

Implementation of the VLDB ini-tiative takes place in consultationwith stakeholders who manageimportant research data. As part ofthe pilot of phase 1, the followingorganisations have been broughton board: the Human Sciences Research Council (HSRC), theSquare Kilometre Array (SKA), the South African AstronomicalObservatory (SAAO), and theSouth African Environmental Observation Network (SAEON).

Once the data repository has beencommissioned and tested, the second phase will commence during the 2011/2012 financialyear. Sithole explains, “All otherstakeholders with important research data will be encouragedto participate.” The storage will bea fully-redundant, reliable systemwith capacity to spare, and will bereplicated at two sites: Cape Townand Pretoria. – Biffy van Rooyen

“The DST’s ICT intervention aimsfor unlimited processing power,bandwidth and storage. We aregradually becominga destination of choice for highperformance com-puting, attractingsome of the bestminds in cyber-infrastructure.”– Mrs Naledi Pandor,Minister of Science and Technology

Enquiries:Dr Happy Sitholehsithole@csir.co.za

| 35 |

IMAGINE THERE WAS A WAY of intervening in a potential elec-tricity cable theft before the crimehappened, or a way of knowingwell in advance that a mainswater pipe in a suburb was aboutto burst? Imagine that traffic lightsknew you were the only car driv-ing through some intersection ona late Saturday night and helpedyou on your way?

The types of technology chal-lenges associated with the abovescenarios are what engineers andcomputer scientists at the CSIR’snewly-created area on AdvancedSensor Networks (ASN) are inte-rested in solving.

The area was established in Sep-tember 2010. However, John Isaac,ASN’s lead, lets on that the estab-lishment of a research group wasmooted as early as 2008. “We initially planned to start a joint research group with the Universityof Pretoria (UP) under the digi-tal@SERA collaboration, but thisdid not happen quite as we envis-aged. Instead, there is currently an ASN initiative at UP and one at the CSIR that are workingclosely together with various jointactivities,” he explains.

In preparation for the joint initia-tive on ASN, Prof Ian Akyildiz ofthe Georgia Institute of Technologywas appointed as an HonoraryProfessor at UP’s Department ofElectrical, Electronic and Com-puter Engineering.

“In Akyildiz, we have access to aworld authority in ASN who hasreally helped define the field. He isvery excited and very determinedto work with us,” says Isaac. “Hehas made his resources and hisworldwide network available tous. We have embraced him as partof the fabric here at the CSIR,” hequips.

What are advanced sensor networks?In his 2002 article, which hasbeen cited over 12 000 times, Akyildiz explains that a sensornetwork is composed of a largenumber of sensor nodes that aredensely deployed either inside thephenomenon or very close to it.Cheap, smart devices with multi-ple onboard sensors, can be net-worked and deployed in largenumbers, he says. These can pro-vide unprecedented opportunitiesfor instrumenting for monitoringand control of homes, cities andthe environment. According toAkyildiz, the position of the sensornodes need not be engineered or predetermined. This allows ran-dom deployment in inaccessibleterrains or disaster relief opera-tions.

Isaac agrees and emphasises thatan important aspect of advancedsensor networks is that the sen-sors are ‘smart’ (i.e. have compu-tational capabilities to make localdecisions/actions) and the net-work technologies connectingthese sensors also have some special features that in combina-tion allow for the intelligence dis-tributed in the sensor nodes to beutilised to cooperatively carry outtasks even without human inter-vention. Thus, advanced sensornetworks allows smart systems tobe created that can sense, decide

and act autonomously or withsome added human direction.

Building an industryOne focus of the ASN initiative,says Isaac, is on the uses of ASNfor electrical utilities for monitor-ing, control and protection includ-ing in extreme and hazardousenvironments. Other areas of in-terest are underground, underwa-ter and multimedia sensornetworks with applications rang-ing from agriculture to mining toenvironmental monitoring.Isaac says that the main aim of the ASN initiative is to give theeconomy a boost, “We work in anarea where we can stimulate in-dustry.” He notes, however, thatASN is still an emerging concept in the country, “We find that, atthe moment, the market is notthat well-educated on what is possible with advanced sensors.”

He states that to stimulate theeconomy, the ASN initiative willplay a strategic role. “We want toget a national conversation goingaround the many possibilities thatASN brings. All of this means thatwe have to talk to a whole lot ofsmall to medium-sized enterprisesand technology vendors. We haveto bring in the universities and cre-ate R&D backing capacity at uni-versities and the CSIR in this field.”Isaac believes South Africa is well-poised to build its own industrialcapabilities around advanced

sensor networks. “South Africa isnot that far off to the rest of the developed world,” he says. “Wehave the advantage that we arestill building our infrastructure,thus while we are fixing waterpipes and installing cables, we canalso embed advanced sensors,” he adds.

One of the most exciting fieldsIsaac states that ASN is an excitingfield that opens up many opportu-nities for collaboration between in-dustry, academia and the CSIR.“This is one of those attractiveR&D areas with multidisciplinaryimpact where it should be fairlyeasy to get various units withinthe CSIR to work together,” hestates. “There is something foreveryone in ASN. For those whowant to publish, there is a lot of research to be done. For industryand others with a profit motive,there are many applications andsolutions that can be built and exploited for profit,” he says.“When all is said, the most important thing is that we have an amazing opportunity to createan industry,” Isaac concludes. – Bandile Sikwane

“This is one of those attractive R&D areas with multidisciplinary impact where it should be fairlyeasy to get various units within the CSIR to work together.”

Extra sensory perceptionAdvanced sensor networks can change the way we interact with our world, says John Isaac,lead member of this emerging innovation area at the CSIR.

Enquiries:John Isaacsisaac@csir.co.za

ICT and the Economy

LIVE TELEVISION broadcastingover the public Internet is not anew idea, but it is one that hasbeen almost impossible to accom-plish in developing countries upuntil now. That is because differ-ent areas and people have accessto different bandwidths, makingthe quality of the broadcast feedunstable and irritating to watch.

In developing countries access tobroadband services are few andfar between. Researchers at theCSIR, University of Cape Town andEast Coast Access have come upwith a solution for this problemthat will soon be put through itspaces in a pilot study by a localradio station. It is called theARTIST project, short for AdaptiveReal-Time Internet StreamingTechnology, where local and inter-national patents on the technologyhave been filed.

The competitive edgeDr Keith Ferguson of the CSIR’sReal-Time Video Coding research

group explains: “Our invention al-lows for anyone and anywhere toset up an online television stationof their own, derive revenue andcreate job opportunities from it.What gives our technology thecompetitive edge is the fact thatthe quality of the image is adapt-able to the quality of the infra-structure, making it ideal for lowbit-rate situations such as ismostly experienced in developingcountries.”

Imagine watching a live broadcastof a world-altering, breaking newsstory; or a sports game; or a musicshow; without the constant buffer-ing or delayed transmission andwhile only having access to a slowInternet connection. Ferguson isadamant: “With live broadcasts,buffering and delays are simplynot allowed!”

He mentions that they want tobroadcast African creativity withthis technology. “We don’t fullyknow yet what can be done withit,” he says, referring to the factthat, as engineers, they cannot

even begin to imagine the creativespectrum thereof. “The possibilitiesare endless, and we cannot wait toput it into the hands of communi-ties and see how creative they canbe.”

A radio community as the pilotThe first community that will beparticipating in ARTIST’s pilot is a local commercial radio stationwith an audience of just over 1.4million listeners. The station isbased in Johannesburg.

“For us it was important to pilotthe technology with someonewhere we would be able to bephysically present to sort out anyglitches. The location of the stationand community demographic fitted the bill perfectly.”

According to Ferguson, the pur-pose of the pilot is not only to testthe technical ability of the techno-logy in a real-world situation, butto also test its commercial viability

Internet television cOmmunity

– live broadcasting for the masses

WHO IS INVOLVED?

In 2009, the CSIR’sreal-time videocoding researchgroup received aninvestment fromthe InnovationFund at the Department of Science and Tech-nology to form aconsortium with aDurban-basedBlack EconomicEmpowermentcompany, EastCoast Access, andthe University ofCape Town.

| 36 |

and possible commercial models –that is, whether advertisers will be willing to use this medium todeliver their message.

“Through the technology the radiostation is given an opportunity tobroaden the scope of its website byattracting more of their listeners.New, innovative ways of onlineadvertising then becomes avai-lable. It also opens up branding opportunities for companies or organisations.

The pilot radio station, for in-stance, is planning to actively promote its listeners to also become viewers and to change its style of presentation to be morevisually interesting through thelive video streaming on its web-site,” he comments.

Peer-to-peer video conferencing protocolstandardisationLinked to the ARTIST project is another from the same group of

researchers that involves high-end,high-definition live video confer-encing over the Internet. “Imaginevideo conferencing with severalpeople at the same time, all inhigh-definition,” says Ferguson.

With this peer-to-peer protocolproject, the group aims to build aninternational profile by contribut-ing to the standardisation of peer-to-peer video conferencing-relatedprotocols. They are working withBerlin-based Fraunhofer’s Insti-tute for Telecommunications, theHeinrich Hertz Institute.

Catering for the lower end of the marketA third project revolves aroundlow bit-rate video coding for appli-cations in developing countries.“Developed countries,” says Fergu-son, “are not primarily interestedin low-definition coding. Theymainly concentrate on improvingthe envelope for high-definition.The reality in the developingworld, however, is limited Internet

communication infrastructurewhich will still be around for manyyears to come.”

The aim of this project is to extendthe lower range of the interna-tional video-coding standards toenhance the picture quality at thelow bit-rate end of the market.

Ferguson explains: “The quality ofyour Internet connectivity changesas you move around on a mobiledevice or as more people share thesame bandwidth and so on. Withthis project, we aim to adaptivelymaximise the picture quality forthe given bit-rate that your clientdevice experiences at any givenmoment in time.”

He says that if one can track theups and downs of their connecti-vity then they can encode thevideo to match the feed that youreceive. “You need to be able tomove in and out of different band-widths all the time while still pro-viding the best perceived viewingquality; and this is what we aim toachieve.”

Funding to commercialiseFor now, all eyes are on theARTIST pilot project with its radiostation pilot. If the technologyproves to be successful and thecommercial models viable, theteam will need funding to take itfurther into the market.

“The world is our oyster with thistechnology. We have already hadmassive interest from various‘communities’ who want to use it.Hopefully, it will bring live, qualitybroadcasts into the homes and offices of millions of people in developing countries and, at thesame time, take on the mobile developed world,” Ferguson concludes. – Petro Lowies

What did the researchand development ofARTIST entail?To broadcast live video over the Internet requiresdesigning massively scalable systems that cancope with hundreds ofthousands of simultaneousviewers and deliver appro-priate advertising to eachunique viewer.

The group’s research, development and commer-cialisation agenda spannedthe modelling of networkbehaviour under congestedconditions, video encodingand optimisation tech-niques, massively scalablearchitectures for commer-cial-grade video deliveryand built-in new media advertising models.

The CSIR’s Real-Time Video Codingresearch group includes, from left,Dhiren Seetharam, Nakato Kakande,Louis Joubert, Dr Keith Ferguson,Veronica Sentongo and Ralf Globisch.Absent from the picture is MardeHelbig.

Enquiries:Dr Keith Fergusonkferguson@csir.co.za

| 37 |

ICT and the Economy

“Phones are inherently social devices and the industry is just beginning to discover what’s possible,” said Facebook bossMark Zuckerberg at the MobileWorld Congress in Barcelona,Spain in February 2011. He was responding to rumours that thecompany is planning to launch itsown mobile phone, and made itclear that they would rather investin the development of mobile ap-plications (m-apps) than physicalhandsets.

While social networking is justone type of m-app that has al-ready taken the world by storm,the fact remains that mobilephones have the ability to connectthe unconnected to the WorldWide Web and to do so in wayswe can only begin to imagine.

A technology boom in the making“The m-apps environment is at anearly stage of a profoundly impor-

tant movement toward establish-ing the next dominant global in-formation platform, and especiallyin the developing world,” says BenZaaiman, programme leader ofthe CSIR Meraka Institute-ledmLab. If statistics are anything togo by, then he is right on themoney.

There are approximately 450 mil-lion mobile subscribers in Africa,which represents a market pene-tration of 50%. The World Bank(WB) estimates that in this region,a further 300 million people willhave access to mobile phoneswithin the next five years. InSouth Africa alone, the marketpenetration for mobile phones already stands at 100% of house-holds. Nine out of every 10 phonelines on the continent are mobile(as opposed to landlines).

“There are many more mobilephone users in Africa than com-puter users,” says Zaaiman. “Measure the above figuresagainst the South African market

penetration percentages for usersof personal computers (6%), andthose with broadband access (4%),of which only 0.5% use ADSLlines. It stands to reason that theprimary opportunities for con-sumption of information are nowthrough mobile phones, and thatthe way organisations communi-cate with their customers shouldbe changing too.”

mLabs to foster m-appsand socio-economic developmentThis is why the WB through infoDev – the Government of Finland and mobile phone manufacturer Nokia – is setting up seven mLabs in different partsof the developing world. Four ofthese mLabs are earmarked forAfrica, two for South-East Asiaand one for Eastern Europe. About $20 million has been ear-marked for the programme.

In Africa, the East African mLabhas been set up at the iHub inNairobi, Kenya. The southernAfrican mLab’s hub is based atthe Innovation Hub in Pretoria,South Africa, with ‘spokes’ (or regional offices) planned for different countries within the region. The first satellite has beenidentified in Cape Town, and itsmanager has already been re-cruited. The Southern AfricanmLab is hosted by a consortium of the CSIR, The Innovation Hub,Innovation Lab and UnganaAfrika.

“The WB is all about socio-eco-nomic development, hence this investment,” says Laurens Cloete,acting Executive Director of theCSIR Meraka Institute. “They believe that the people in the developing world are able to comeup with innovative ways in whichto tackle their own specific problems. The aim is to have atwo-way flow of knowledge; byempowering them to become entrepreneurs either through

mLabs: Empowering through locally-developed

mLab = Regional Mobile Applications Laboratory

| 38 |

developing m-apps themselves orby using m-apps developed forthem, faster economical develop-ment can take place.”

Nokia’s role as a technical partnerWhere infoDev acts as the imple-menter, Nokia acts as the technicalpartner. Cloete explains: “Ensuringthat content and application deve-lopers have access to the latestmobile technologies and trainingprogrammes, Nokia gives ‘in-kind’ contributions such as bring-

ing out experts to provide training or giving access to theirhandsets on which entrepre-neurs can test their newly developed m-apps.”

According to Zaaiman, Nokia is by far the world’s dominantsupplier of mobile phone hand-sets. “Yet these are predomi-nantly feature phones, notsmart phones,” he says. “And in the developing world featurephones are the ones that areused overwhelmingly, eventhough the world is moving to-wards the use of smart phones.”

At the AITEC East African ICTSummit, which was held inSeptember 2010 in Nairobi, Jussi Hinkkanen, Nokia’s VicePresident for Government Af-fairs and Business Environmentin the Middle East and Africa,said: “Mobile application andcontent will reach even themost disadvantaged populationand provide new sources ofwealth and innovation for

entrepreneurs of the southernAfrican region.”

What the mLabs will doWhile the CSIR-led mLab is still in an establishment phase, it hasalready attracted support fromseveral industry players to assistwith its continuation once theseed funding from infoDev and itsFinnish partners has served itspurpose.

The mLabs will provide:• technical and business trainingand mentorship to potential m-apps entrepreneurs and developers;

• test laboratories; • advocacy of mobile usage andapplications; and

• structured channels in the marketplace between industryplayers (such as networkproviders and sms gateways).

“Apart from providing potential m-apps developers and innovatorswith all the support and facilities

mobile applications

Enquiries:Laurens Cloetelcloete@csir.co.zaBen Zaaimanben@dcisio.com

they could possibly need, we seeourselves as also being an imple-mentation partner for NGOs andother public institutions that areparticularly concerned with servicedelivery issues,” says Zaaiman.

The mLabs are all expected to pro-vide at least eight commercial ap-plications within two years. – Petro Lowies

infoDev is a global development financingprogramme among international develop-ment agencies, coordi-nated and served by anexpert secretariathoused in the Global ICT Department (GICT)of the World Bank.

Laurens Cloete, acting Executive Director of the CSIR Meraka Institute, and Ben Zaaiman, manager of the CSIR-led mLab.

| 39 |

“Mobile applications andcontent will reach even the most disadvantagedpopulation and providenew sources of wealth and innovation for entrepreneurs of thesouthern African region.”

ICT and the Economy

The SAP Meraka Unit for Tech-nology Development (UTD), a unique South African public-private partnership between government, a multinational inthe field of collaborative businesssoftware and the CSIR Meraka Institute, occupies a singular research, development and innovation spot in South Africa’sinformation and communicationstechnology (ICT) landscape. Theunit focuses on technologies foremerging economies, specificallyInternet applications and services.

SAP MERAKA UTD was the resultof foreign direct investment bySAP in 2007, which was matchedthrough investment by the Depart-ment of Science and Technology(DST) and the CSIR Meraka Insti-tute.

The big pictureDanie Kok (Director: SAP ResearchIA&S and co-manager: SAP Me-raka UTD) is clear on what theSAP Meraka UTD envisages, “We believe that we have a role toplay by contributing to the build-ing of local ICT research capacity.This is done by evolving the unitto a world-class research and de-velopment laboratory in the fieldof technologies for emergingeconomies.” Technologies foremerging economies – explainsProf Jan Eloff (Research director forSAP Meraka UTD) – can be eithernew technologies or adapted tech-

nologies (taken from a first worldcontext) and applied to businessservices for very small enterprises,for energy and for mobile healthmonitoring. “We aim to delivertangible innovations,” Eloff asserts.

He believes that the SAP MerakaUTD owes much of its success toits understanding of the chal-lenges faced in emergingeconomies. Small and mediumenterprises (SMEs) hold the key toinfluencing socio-economic condi-tions in a positive way. Achievingeconomic sustainability remains avitally important factor for SMEs.Equally important is the impera-tive to provide access for aspiringentrepreneurs to the informalbusiness sector.

Implicit in this understanding isthe recognition of two importantdrivers: Co-innovation and busi-ness impact. “Our joint research

undertakings must deliver

value in terms of business,” Eloffconfirms. “Based on our under-standing of this principle and ofthe challenges in emergingeconomies, and in line with ourstakeholder expectations, we haveadapted our research agenda.”

Eloff sets these out: Research anddevelopment of new ICT solutionsfor emerging economies; measure-ment and validation of social andeconomic impact; and investmentin methodologies, technologiesand techniques for emerging economics.

This research agenda has in turnfound expression in four directedresearch projects.

ICT in the service of emerging economies

Fast facts on HCD@SAP Meraka UTD in 201112 full-time students

20 interns and students

First completed doctorate:Dr Marek Zielinski

| 40 |

(Front) Danie Kok (Director:SAP Research IA&S and co-manager SAP Meraka UTD) and Jason Chuang(Middle) Prof Ernest Ngassm and Elmarie Venter(Back) Dr Marek Zielinski, Prof Jan Eloff (Research Director: SAP Meraka UTD) and Dr Danie Smit

e-EnergyRural and low-income householdsin South Africa are characterisedby usage of a mix of energysources (e.g. wood, electricity, gas and paraffin) for their dailyheating, cooking and lightingneeds. The e-Energy project aims to understand energy usage patterns in households in KwaMbonambi (northernKwaZulu-Natal), and will includequantifying usage of non-electri-fied energy sources. A mobilephone-based software consolegathers information about energyusage in households. It also offersa tool for disseminating exemplaryincentives to foster change in energy usage patterns of house-holds.

Looking aheadKok and Eloff are keen for SAPMeraka UTD to continue to de-velop innovative applications foremerging and fast-growing mar-kets, as well as approaches andtechnologies for delivering innova-tive applications (cloud, mobile, ondemand and on device).“As a local research unit, the SAPMeraka UTD contributions areunique and valued by SAP,” Kokconcludes. “We believe that wecan continue to deliver novel solu-tions for emerging markets, whichcould be useful in the developedworld.” – Biffy van Rooyen

| 41 |

RusticaThe first of these projects, Rustica,seeks to overcome barriers inhibit-ing rural entrepreneurship and thelack of access to mainstream orglobal supply chains and markets.This is done by introducing a largeset of ICT solutions to the rural Living Lab community ofKgautswane in the SekhukhuneDistrict Municipality in Limpopo. Kgautswane consists of 19 villageswith a population of about120 000, served by some 130spaza shops. Rustica will intro-duce a system of a virtual buyingco-operative (supported by theCSIR’s network of Infopreneurs®),an increased product range andadditional suppliers. The projectwill measure the effect of this intervention on the social and economic development in thiscommunity.

OvertureVery small enterprises (VSEs) offewer than 20 people face severechallenges in terms of profitability,sustainability and the ultimatesuccess of their businesses. To address the challenges of access tocritical business services, the Over-ture project has devised a mobilebusiness applications platform.Services are delivered via a mobilephone interacting with a collabo-rative multi-tenant back-end envi-ronment. The back-end is hostedby an intermediary such as a mo-bile telecommunications provider.Vodacom/Vodafone partnered inthis project.

IsagoTo broaden the scope of the mobilebusiness solutions to VSEs, theproject, Isago, aims to develop andevaluate a next generation plat-form. This will offer generic busi-ness services to support theireconomic sustainability. The inno-vation lies in the development of aprototype product catalogue formobile phones as well as an algo-rithm for the intelligent retrieval of products. The latter takes intoaccount factors such as the user’scontext, location, purchase historyand preferences and limits theamount of data sent to the phone.A future development will be analgorithm for intelligent presenta-tion of product items, taking thecharacteristics of individual mobilephones into account.

Enquiries:Prof Jan Eloffjan.eloff@sap.com

Spaza shops are small convenience stores, operating out of people’s homes or backyards and providing a vital service to people living in under-resourced communities.

SAP/Meraka UTD collaborates closely with theCSIR’s Living Labs research efforts, led by ProfMarlien Herselman. The Living Labs approachensures participation by all parties (includingbeneficiaries) in all phases of the project.

ICT and the Economy

THE INPUT DATA to the web-based calculator needed are fuelconsumption; electricity cost orconsumption; employee air travel;airborne, rail cargo weights anddistances; and inventory of refrig-erants. The calculator will help theenterprise estimate emissionsfrom operations and take neces-

sary remedial action if required.This methodology for calculatingemissions emanates from theusage of transportation fleet and acarbon footprint from enterpriseactivities in general. The two maingreenhouse gases deemed tocause global warming have beenidentified as CO2 and CH4.

Other greenhouse gases includeN2O, hydrofluorocarbons, perfluo-rocarbons and SF6, although theseoccur to a lesser extent than theformer two.

Dr Njabulo Siyakatshana, projectleader, says that the aim of thisnovel web-based emissions calcu-

Novel web-based calculator to estimate emissions and

general carbon footprintThe CSIR has designed a web-based calculator to quantify carbon emissions.

| 42 |

Dr Njabulo Siyakatshana

lator is for the client to be able toaccount for its emissions. “Theclient wants to have a databasewhere it can quantify its carbonemissions,” he says. “The objectiveis to encourage emissions account-ability in state-owned enterprises.”

He adds that the aim of the clientis to look at its strategy and seewhere it can reduce greenhousegas emissions and carbon foot-print.

Siyakatshana and his team did notdevelop any new technology. Thisapplication, however, is unique tothe client’s needs. “The clientwanted to quantify the emissionsreleased from its electricity con-sumption, vehicle fleet usage andother forms of freight transporta-tion such as airborne, rail androad,” he says. “We also included abalance of refrigerants to quantifyemissions resulting from air condi-tioning.”

Tools usedThis, notes Siyakatshana, was adesktop study; no field work wascarried out. He says that his teamcollected data from the UnitedStates Environmental ProtectionAgency, the Society for Automo-tive Engineering and the IPPC because “these organisations already have comprehensive andopenly-available databases onemission factors”.

“With this kind of data, we couldreach our objectives by developingthis user-friendly application,” hesays.

Siyakatshana adds, “Although thisapplication is specifically designedfor the client, it is customisable forother sectors because the method-ology would be similar.”

Impact“This will have an environmentalimpact on both the public and theprivate sectors,” he says, addingthat neither of these sectors hasdeveloped methodologies to accu-rately quantify their carbon emis-sions. “There will come a timewhere South Africa – as a signa-tory of the Kyoto Protocol – willhave to account for its carbonemissions.”

Meanwhile, emissions from theenterprise activity in considerationwere categorised as follows:

Vehicular emissionsThe main emissions from roadtransport are mainly carbon dio-xide (CO2), carbon monoxide (CO),nitrogen oxides (NOx), hydrocar-bons (HC), methane (CH4) and par-ticulate matter (PM). These are inaddition to heavy metals release,mainly lead (Pb) and sulphur ox-ides (SOx) and a host of other un-

regulated emissions such as polyaromatic hydrocarbons and per-sistent organic pollutants. Vehiclesconsidered in the study range frompassenger cars, light duty, heavyduty, motorcycles and forklifts.Input data for these calculationsare petrol or diesel consumption inlitres.

Electricity-based emissionsAll electricity-based operations release carbon dioxide and othergreenhouse gases during their lifecycle. These emissions can bedirect, that is, arising from firingthe power plant, or indirect, thatis, arising from other phases suchas mining and transportation ofthe fuel during its life cycle. Themain greenhouse gases tabulatedare CO2, CH4, and N2O.

Air travel emissions andairborne cargoThe calculation of emissions fromflights is an evolving and compli-cated area, and different calcula-tors available result in differentvalues. One factor that has a largeinfluence on the calculation is theRadiative Forcing Index (RFI). Thisincorporates the global warmingeffects planes have because of thealtitude at which they emit. An average value tabulated by theInter-Governmental Panel on Climate Change (IPCC) was used.

Rail freightThese are normally tabulated asCO2 equivalents. The Europeanstandard of CO2 25 g/ton-km areactually on a downward trendfrom a high of CO2 35 g/ton-km in1990. However, these figures aremore significant for enterprises operating chiefly on rail cargotransport.

RefrigerantsAir-conditioning equipment is asource of potent greenhouse gasesin the form of chloro-fluoro car-bons. They have global warmingpotentials thousands of timesgreater than CO2 although they arenot released in comparable quanti-ties. These are released chiefly byleakage during charging, opera-tion or servicing. An up-to-dateinventory and mass balance helptrack the leakages and quantifythe emissions. – Mzimasi Gcukumana

Enquiries:Dr Njabulo Siyakatshanansiyakatshana@csir.co.za

| 43 |

| 44 |

Digital DoorwayTM –Opening the doors to ICTsTHE DIGITAL DOORWAY (DD)initiative was introduced five yearsago by the Department of Scienceand Technology (DST) and the CSIRas a way to improve computer liter-acy and access to information.

Underpinning the project is the concept of people’s inherent abilityto teach themselves and gain information with minimal externalintervention. For this to happen,computers must be easily accessibleto potential learners (children andadults) in an environment con-ducive to experimentation.

The robust DD can be placed wherethey are most accessible to users.There are currently 220 systems inSouth Africa.

The CSIR is also collaborating withMonash University of Melbourne,Australia, and UNICEF Pacific toreach remote communities in theseareas and help overcome the digitaldivide.

The DD comes in different configu-rations: single seater; three seater; for people with disabilities; and the container Digital DoorwayTM.

The latter is run on solar energy and is rugged enough to be placedin exposed locations. – Bonang Tselane

“Chatting our wayto better maths”DR MATH is a mobile tutoringservice which allows mathematicslearners to use the MXit mobile instant messaging service on theircell phones to access tutors whoprovide real-time support and assistance with mathematicshomework and revision. This project was initiated in 2007; there have been 17 000 registeredusers to date.

When using Dr Math, learners areable to have a one-to-one ‘conver-sation’ with a tutor who supportsthem in solving mathematicalqueries (e.g. maths homework) in real-time. The service can belikened to a text-based call centre,where individual learners are as-signed to tutors, based on tutoravailability.

Tutors are people who have astrong mathematical backgroundand who are trained to use DrMath. They are required to under-stand how to use the medium and conform to a strict code ofconduct. The tutors are managedin a similar way to call centreagents and are available Sundaysto Thursdays between 14:00 and20:00. There are currently 110 volunteers from around the country – most are engineeringstudents from the University ofPretoria.

The CSIR has, however, over thepast six months developed ageneric set of tools that enablesanyone to set up a similar serviceto Dr Math. The toolset is knownas ‘C3TO’ (Chat Call Centre TutorOnline), which currently uses theMXit platform as a channel, butcan utilise other platforms thatsupport Google Talk.

CommercialisationThe CSIR has a long-standing rela-tionship with the Department ofBasic Education, a major roleplayer in the successful implemen-tation of the Dr Math initiative.

The CSIR is also working on licensing Dr Math to an inde-pendent organisation that will be capable of offering the servicenationally (initially and thereafterinternationally) and grow the initiative to achieve impact and deliver on stakeholders’ expectations. – Bonang Tselane

recapping

Enquiries: Nicki Koorbanallynkoorbanally@csir.co.za

Enquiries: Nicki Koorbanallynkoorbanally@csir.co.za

Visit: ht

tp://

www.

csir.c

o.za/

publi

catio

ns.ht

mlfor p

revio

us is

sues of

the S

cienc

eSco

pe

| 45 |

Making ICTs available to persons with disabilitiesThe National Accessibility Programme (NAP) was initiated in 2004 as a five-year research programme to address the marginalisation ofpersons with disabilities from the mainstreamsociety and economy. This project also sets out toensure that persons with disabilities participateand are included at all levels of society throughthe use of information and communicationstechnologies (ICT).

Key assets in the NAP suite include:• An accessible web source code that enablesassistive technologies such as screen readers

• A website/portal using the source code and setup for persons with disabilities

• Technical developments such as GNApp (anopen source augmentative and alternativecommunication framework) and the Notetaker(a computer for the blind).

NAP looks to service the new Ministry forWomen, Youth, Children and Persons with Disabilities, as well as other government departments.

In order for NAP to achieve maximum potentialand impact, it must be transferred to an appro-priate implementation vehicle that will be re-sponsible for operationalising an independentand sustainable NAP and its associated interven-tions nationally and into Africa.– Bonang Tselane

Seeing the worldthrough differenteyesThe dream to have access to elec-tronic documents and informationand be part of the global move to-wards accessing the Internetthrough modern communicationsand computing tools, has now be-come reality and will soon beavailable to the average visually-impaired consumer.

This is thanks to the work of theCSIR in the development of theNotetaker. The Notetaker – a ‘computer for the Blind’ – is an indigenous technology focusingon low cost, affordability andvalue benefits. It is aimed at improving accessibility by main-streaming persons with disabilitiesinto society through the use of information and communicationstechnology (ICT). This technologywas designed and developed bysenior technologist, Willem vander Walt, who is blind.

In 2009, the World Health Organi-zation (WHO) estimated that therewere about 314 million peopleworldwide living with visual impairment. Of those, 45 millionwere blind, 90% of whom wereliving in low-income countries.This meant that a lot of blind people are in disadvantaged situa-tions due to a lack in resources.

The ‘computer for the Blind’ is afirst of its kind in South Africa because it is a customised solutionthat provides support for multipleindigenous languages, currentlyincluding English, Afrikaans andIsiZulu. Notetaker also has the potential to make immediate impact in the education and employment sectors.

The research and developmentstage of this technology is nowcomplete and it has reached thecommercialisation stage. The CSIR is currently working with Micro Link SA, an implementationpartner to take the Notetaker tothe market. Micro Link PC spe-cialises in the supply of computerhardware, software and other specialised equipment not only tosupport people with disabilities,but also to enable professionals tomaximise their output within theirworkplace.

In addition to Micro Link, the CSIRis exploring other partnerships forproduct development and testingas well as for marketing, supportand maintenance of the notetaker.– Bonang Tselane

Enquiries: Hina Patelhpatel@csir.co.za

Enquiries: Nicki Koorbanallynkoorbanally@csir.co.za

S

Computational Platforms

IBM e1350 Cluster – 640 CPUs at 2.5 teraflops measured performance.IBM Blue Gene-P – 11.5 teraflops measured performance.IBM P690 SMP – 166 gigaflops peak performance.

Sun SPARC Enterprise M9000 server with 64 SPARC64 VII quad-coreprocessors, and a cluster of four Sun Blade 6048 Modular Systems. The cluster houses a total of 192 nodes based on the next-generation IntelXeon E5450 processor (Nehalem) – 27 teraflops total peak performance.

Storage solution based on ten AMD Opteron-powered Sun Fire X4540Open Storage servers, providing 480 terabytes of data with the Lustre parallel file system for extreme I/O performance and reliability.

Range of clusters with other computational components:GPGPU – General Purpose computing on Graphics Processing UnitsFPGA – Field-Programmable Gate Arrays.

Research and Education

The CHPC is part of South Africa’s national cyberinfrastructure intervention: Supported by the Department of Science and Technology(DST) and hosted by the Council for Scientific and Industrial Research(CSIR). Other complementary national cyberinfrastructure includes: South African National Research Network (SANReN) – provision of highspeed, highbandwidth connectivity, and Very Large Data Sets (VLDS)– effective curation of a variety of notably large data sets.

Advancement of scientific boundaries by enabling world-class research through promoting and facilitating the use of computationaltechnologies and techniques amongst researchers.

Fostering of innovation through effective partnership for training a new generation of skilled researchers in areas underpinned by high performance computing and data curation, particularly those of strategic national and continental importance. Promoting private public partnership and the utilisation of the high performance computingby the commercial sector.

The CHPC invites all researchers to take advantage of its cyberinfrastructure.

The CHPC seeks joint research and HCD initiatives with its stakeholders.

For more informationwww.chpc.ac.za

Lore

tta S

teyn

Gra

phic

Desig

n St

udio