Post on 26-Apr-2018
1
iMinds Technology
and Expertise for
IoT Large Scale Pilots
for Wearables (LSP3), Cities (LSP4),
Ageing Well (LSP1) & Vehicles (LSP5)
Stefan Van Baelen
Research Coordinator
Stefan.VanBaelen@iminds.be
© EPoSS 2015 EPoSS Brokerage Workshop on IoT Large Scale Pilots, London, 2015-06-26
3
47 mio
380+
1,000+
850+ TOTAL REVENUE RESEARCHERS located at 5 Universities
(KU Leuven, UAntwerpen, UGent, UHasselt, VUB)
START-UPS from the iMinds incubation programs
PARTNERS in the iMinds ecosystem
PROJECTS with industry (FP7/H2020, EIT,
AAL, ARTEMIS/ENIAC/ECSEL, EUREKA
ITEA/Celtic-Plus/CATRENE, National)
80+
1,000+ PUBLICATIONS on a yearly basis
Flanders’ Digital Strategic Research Center
& Incubator
4
Innovation Toolbox & Strategic Programs
5+ y …1y Time-to-market
Strategic
Research
Incubation &
Venturing
Applied
Research
Test, Validate,
Co-create
• Strategic
Research
• Research
Excellence
• Strategic
Partnerships
• Training &
Coaching
• Financing
• Facilities &
Networking
• Internationali
-sation
• Demand
Driven
• Inter-
Disciplinary
• Cooperative
• PoCs
• iMinds
Living Labs
• iLab.t:
Technical
Test-Beds
Digital Security
Internet of Things
Big Data Analytics
High-Tech
Visualisation
National &
International
Collaborative R&D
City of Things
City LivingLab
in Antwerpen
(BEL)
Incubation,
iStart, iBoot
5
iMinds IoT High-Impact Program From Lower to Higher Market Readiness Levels
iMinds IoT Research Centre
Smart Cities: City of Things
Living Lab
Residential: Home lab
Manufacturing: Mobile lab
IoT Bootcamp
IoT opportunity recognition workshops
IoT Training
Ma
rket A
doption R
eadin
ess
Ghent Lab Leuven Lab
6
CONNECTS
SMART CITIZENS
WITH
AND CREATES THE
LARGEST LIVING LAB
OFFERED TO
TECH DEVELOPERS & COMPANIES
DEVICES
City of Things
10
Tackling 4 Challenges for IoT Market Adoption
Ensure
Wireless
performance
Efficient
IoT Data
Analytics
Effortless
deployment
of IoT solutions
Secure IoT Architectures
1 2 3
4
11
iMinds Technology Roadmap
Ensure
Wireless
performance
Flexible Radio Platform
Efficient
IoT Data
Analytics
Distributed reasoning on distributed IoT data
Black/Grey/White-box adaptive system modeling and control
Flexible protocol development
Reliability in dense and harsh environment
Distributed Network Intelligence
Effortless
deployment
of IoT
solutions Effortless adaptive distribution of IoT Intelligence
Scalable collection and processing of IoT Data
Plug-n-Play Sensor Integration
Unsupervised event and anomaly detection in (real-time) IoT data streams
Self-managed, cognitive IoT architectures
Predictive/corrective maintenance in industrial setting
Secure
IoT
Architectures
Advanced support for policy enforcement and management
Cryptography for IoT
End‐to‐End Security Architectures
Hard ‐and software primitives for secure IoT ...
1
2
3
4
12
iMinds IoT Technology Assets
1. LooCI: Interoperability for heterogeneous
embedded systems
2. μPnP: Plug and play addressing of IoT peripherals
3. CUBE: Deterministic wireless networking
4. Atto Cell: High-bitrate, low latency wireless cells
5. AIOLOS: Computational offloading
6. DYAMAND: DYnamic, Adaptive MAnagement of
Networks and Devices
7. DiANNe: Distributable neural networks
8. PRESENT: Ultra-Lightweight Encryption
13
The iMinds Living Labs toolbox 5 unique assets
A toolbox for any project type: ICON, Living Lab, CIP, FP7, …
Panel Management We’ll find and motivate your test-users
Living Lab Methodology We’ll show you how to set up a living lab project – e.g. with LL Analyser and Data Aggregator
Prototyping & testing We’ll model a rough idea into a usable app for daily life and test it through
Simulate Your Business Co-design of cooperative business model on the fly
European Network of Living Labs
Gateway to 300+ Living Labs
Importance of validated toolbox!
14 © EPoSS 2015 EPoSS Brokerage Workshop on IoT Large Scale Pilots, London, 2015-06-26
Supply side or demand side Supply side
IoT Large Scale Pilot (LSP)
addressed
Focus esp. on Wearables (LSP3) & Cities (LSP4),
also on Ageing Well (LSP1) & Vehicles (LSP5)
Which part of the LSP is
addressed?
SW platform, Architecture, Communication, WSN,
Deployment, Data Analytics, Security, Applications
Description of technology or
solution offered
LooCI, μPnP, Cube, Attocell, AIOLOS, DYAMAND,
DIANNE, PRESENT, IoT Security,
IoT Technical Labs, Home Lab, Mobile Lab,
City Of Things City Living Lab
Description of the proposed
approach and contribution to
the LSP
Technical Contribution, Lab Infrastructure, Deployment,
City Living Lab
Partners already involved
(supply side / demand side)
-
Partners needed
(supply side / demand side)
-
Estimated budget for LSP -
Expected duration of the LSP -
Overview slide
15 15
CONTACT DETAILS Thomas Kallstenius
Director Research & Innovation Strategy iMinds
Vice-chair AIOTI WG7 Wearables
Thomas.Kallstenius@iminds.be
Stefan Van Baelen
Research Coordinator iMinds
Stefan.VanBaelen@iminds.be
16
LooCI Interoperability for heterogeneous embedded systems
• Component-based middleware (execution environment,
component model & event-based publish/subscribe
binding model)
• A reparameterizable component model for networked embedded
systems
• Runtime deployable
components
• A distributed pub/sub
event bus hiding
distribution aspects
from application
components
• Interoperable
implementations on
various platforms
(Contiki, Sun SPOT, OSGi and Android)
Effortless
deployment 1
1
https://distrinet.cs.kuleuven.be/software/looci/
17
μPnP Plug and play addressing of IoT peripherals
• Embed low cost and low-power passive
identifiers in hardware peripherals.
• High level driver language that cleanly
separates peripheral and platform concerns.
• IPv6-based multicast service discovery, driver
installation, usage and management.
• Tiny software stack (< 15KB ROM / 1.5KB RAM)
and compatible with existing peripherals.
Effortless
deployment 1
2
https://distrinet.cs.kuleuven.be/software/looci/wp-content/uploads/flyer.png
18
CUBE Deterministic wireless networks
• The CUBE research track aims to provide reliable,
deterministic wireless network connectivity, where
wireless links behave as if they were deterministic,
configurable (“wired”) links, supporting
heterogeneous applications and devices
Continuous high throughput (HD)
Limited jitter Mbps
100% Reliability
Arrival
us ms s
Point-to-multipoint emergency event
1 event Mbps
100% Reliability
Arrival
us s ms
Example communication streams
Wireless
performance 2
3
https://www.iminds.be/en/news/newsitem_20141204_kickoff-iot
19
Atto Cell High-bitrate, low latency wireless cells
The Atto Cell research track strives towards a
further miniaturization of cells, aiming to create very
high bitrate / low latency / high density networks
Targets:
> 10 Gbit/s/cell
> 1 cell/m2
< 1 user/cell
<10 µsec latency
Wireless
performance 2
4
https://www.iminds.be/en/news/newsitem_20141204_kickoff-iot
20
Generic platform
for runtime
software distribution
OSGi-based platform
AIOLOS provides:
▪Service discovery and binding
▪Fast remote service calls
▪Runtime (re-)deployment
▪Service monitoring
http://aiolos.intec.ugent.be
Data
analytics 3
AIOLOS Computational offloading
5
AIOLOS: Middleware for improving mobile application
performance through cyber foraging
T Verbelen, P Simoens, F De Turck, B Dhoedt - Journal of
Systems and Software, 2012
21
DYAMAND DYnamic, Adaptive MAnagement of Networks and Devices
Data
analytics 3
6
Provides a unified
interface to
many industry
protocols,
including:
... and more http://dyamand.intec.ugent.be
22
DiANNe Distributable Neural Networks
Distributed Artificial Neural NEtworks
Optimize runtime execution of neural networks:
▪ Distribute parts of the neural network among your devices
▪ Train / optimize your neural network structure based on
your devices and their compute capabilities
▪ Prioritize neural network outputs based on latency
requirements
Data
analytics 3
7
https://www.iminds.be/en/news/20150312_news_tim-verbelen-in-the-picture