Enterprise Architecture considerations governing the...

Enterprise Architecture considerations governing the implementation of an Industrial IT strategy at Anglo Platinum

Thursday, June 19, 2008Thursday, June 19, 2008

EA practitioners conference - June 2008

Presented by Johan van TonderManager: Information Management Anglo Platinum

- 2 -Anglo Platinum • Proprietary and ConfidentialEnterprise Architecture considerations governing the implementation of an Industrial IT strategy at Anglo Platinum by J van Tonder June 2008

Agenda for discussionAgenda for discussion

1. Understanding the business objectives of an Industrial IT program

2. Key considerations across the EA stack influencing the design

3. The importance of an effective approach to architecture

4. The IIT applications and technology architecture at Anglo Platinum

5. Key functional MES requirements for the minerals processing industry


An Industrial IT strategy focuses on Integrated Automation and Access to Information to support Decision Making. An Industrial IT strategy focuses on Integrated Automation and Access to Information to support Decision Making.

1. Metallurgical processes are continuous and highly automated. Effective control depends on a constant stream of high quality data.

2. Effective management of large integrated metallurgical complexes requires managers at all levels to coordinate their efforts.

3. A sustainable approach to Operational Performance Monitoring requires all levels of management to work from the same set of reliable information.

4. This enables operational effectiveness and efficiency targets to be agreed, to maximise business value.

5. Often attempts at metrics based management falter because the data gathering process is unsustainable.

6. Globally, systems used by Commercial and Industrial IT are converging


The road to Business and Operational Intelligence cannot be walked in a day… Will your design stand the test of time?

Q208 2009

‘As-Is’

H208 2010 2011

Atedf

‘To-Be’

Past

Ente

rpris

e ar

chite

ctur

eO

rgan

izat

iona

lch

ange

Smelter deploymentConcentrator deployment Refineries deployment

Work completed Current focus Longer term outcomes

Prod

uct

deve

lopm

ent

Agreed standards for all layers of EA for

IIT

Interim release for connectors

Main release: data contextualisation, multiple models &

templatesContracted

operational support

Refineries uses information from

automated sources

Agreed portal strategy in production

Smelter A uses updated reports

IIT plans incorporates

PLC roadmap

Smalter B uses updated reports

Agreed information

architecture for Smelters

Product is the best in MMM

Abnormal situation

management uses

competency centre

Production scheduling

active at refinery

Feed blend optimisation

active Enable asset

optimisation on MES ERP base

Global roll-out

Implemented PI platform for Smelter A

Agreed the design and roles

Phase 1 implemented for

furnaces

Phase 1 implemented for

milling

Concentrator dashboard in Excel, mostly manual data

capturing

Agreed conceptual

MES architecture

Phase 1 supported by management

Cross functional team lead the IIT

program

Competency centre operational

Agreed design for operational

support of IIT First plant uses automated dashboard

Phase 1 functionality across all

Concentrators

Agreed functionality for

product replacement

Agreed functional roadmap for IIT

deployment

All refinery data sources

in MESPilot uses MES

informationAgreed information

strategy


Business process considerations in the Industrial IT worldBusiness process considerations in the Industrial IT world

Human driven processes

– People execute the process steps

Human machine interactions in processes

– People control the process

– Machines execute the process

Automated machine interactions– Machines control the process

– Machines execute the process


Information needs in the Industrial IT worldInformation needs in the Industrial IT world

BRPM Sec Mill Feed Stability – Flowrate

QC AnalysisStrategic decisions

Decisions to control process execution– Geographically disparate management

structure

– Continuous operations

– Functional experts

High degrees of sophistication– Tendency towards “lights out” automation

– Metallurgical processes are multi-variant

– Automates and Artificial Intelligence

Need for expert analysis– Collaborative decision making

– Line execution

– Experts on call

RPMA Merensky Conc Primary Grind - passing 75um [%]QC Analysis

From 10/01/08 06:00 To 10/03/08 22:00 (58.47% of data valid)WARNING LIMIT VIOLATION

10/01/0825/01/08

09/02/0824/02/08

04/0310/03/08

Date / Time

X: 42.70 (40.00); Sigma: 7.13 (1.00); n: 1.

10

20

30

40

50

60

70

80

90

100

RPM

A M

eren

sky

Con

c Pr

imar

y G

rind

- pas

sing

75u

m

[%]

3436404446

Capability(Last Set)Cpl = .40Cpu = .15Cpk = .15

Control Limit ExceededPersistant warning limit exceededWarning limit exceededIn control dataMissing data


Five key models enables a contextualized view of data without compromising integrity or flexibilityFive key models enables a contextualized view of data without compromising integrity or flexibility

Event model

Event 10

Object 101

Process 10

Event 20

Process 20

Product 102

Attribute 1

Attribute 2

Event 30

Object 301

Inventory model

Inventory 10

Asset 10

Product 102

Process 10

Inventory 20

Object 201

Object 202

Attribute 1

Attribute 2

Inventory 30

Object 301

Asset model

Asset 10

Object 101

Object 102

Asset 20

Object 201

Object 202

Attribute 1

Attribute 2

Asset 30

Object 301

Product model

Product 10

Product 101

Product 102

Product 20

Product 101

Product 202

Attribute 1

Attribute 2

Product 30

Object 301

Process model

Process 10

Object 101

Product102

Product 202

Process 20

Object 201

Object 202

Attribute 1

Product 202

Process 30

Object 301

Product context

Process context

Inventory context

Asset context


Technology considerations in the Industrial IT worldTechnology considerations in the Industrial IT world

Often linked to physical plant assets

Safety is a big consideration

“Always on” technology

Speed of data processing

Remote location often implies poor connectivity

Historical context implies “many different standards”


Your operating model will deliver an architecture and resulting product, but is that what you intended?Your operating model will deliver an architecture and resulting product, but is that what you intended?

In 1884, Sarah Winchester, heiress to the Winchester Arms fortune, bought a 6-room house in Santa Clara while still under construction.

She promptly discarded the plan, and kept 22 carpenters full time employed for 36 years!

The house eventually had approximately 160 rooms, with some very unique features:

– 3 elevators

– 47 fireplaces

– Rooms around rooms

– Staircases leading to no-where

– Doors opening to blank walls

– Doors opening to steep drops to the lawn

This was a perfect match for the owner’s needs,

- but she was trying to confine ghosts! www.prairieghosts.com/winchester.html


Operational Intelligence is s sub-set of a sound Information Management strategyOperational Intelligence is s sub-set of a sound Information Management strategy

Support AAplc in their pursuit of

global initiativeswithin the group

• Align with Group IM in support of AAplc shared service objectives in Finance, HR, Supply Chain and SHE

• Reduce the cost of commodity IT services in agreement with business

Supportinternal competency

developmentin pursuit of continuous

improvement• Develop internal competency in

the management of information

• Develop internal competency in enterprise architecture practices

• Pursue a deliberate talent management program

Deliver and support industrial information technology systemsin Anglo Platinum’s

operations• Support operational performance

monitoring by leveraging data contained in base systems

• Develop and support the enabling industrial technology infrastructure in Process division

• Agree an automation strategy in Mining

Enable easier access to

high quality informationacross all levels of management

• Follow a consistent methodology in deploying business projects aimed at delivering better information

• Provide leadership to an information management competency centre in Anglo Platinum, incorporating AAplc group initiatives

Supportthe implementation of sustainable improvements in business process effectiveness• Support business process standardization by applying enterprise

architecture principles in the modelling of business processes• Develop and implement a business process management methodology

capable of supporting controlled changes to established practices• Develop a business process training methodology

Manage expectations through improved relationship managementIntroduce competency centers as a means of improving deliveryImplement an integrated applications management strategy Utilize program management as a means of controlling delivery

Agile delivery of integrated solutions within a predictable time frameUtilize proactive portfolio management to ensure effective implementation of investment decisions

Our

del

iver

ypl

atfo

rmIn

tern

al a

nd e

xter

nal

stak

ehol

der e

ngag

emen

tO

ur v

alue

-add

to A

nglo

Pla

tinum


A sustainable approach to real time information-based decision making in the mineral processing industry requires a layered architectureA sustainable approach to real time information-based decision making in the mineral processing industry requires a layered architecture

Minerals processing

execution (MPX) functions

Minerals processing

execution (MPX) DB

Field devices

Control systems

Minerals processing execution systems

Information Management

systems

Supervisory systems

Programmable logic controller

(PLC)Objects DB

Robo labControl(RLC)

Weighbridge(WB)

Laboratory information

management system (LIMS)

Plant Historian 10 sec Analytical results

Enterprise applications

Operational performance management

(OPM)

Metal accounting (MA)

Enterprise resource planning(ERP)

Metal forecasting (MF)

Processing database

Enterprise data warehouse

Robotic instrument

(RI)

Controller Interface

node

(CIN)

Soft sensor

(SS)

Addressable field device

(AFD)

Weigh bridge

(WB)

Analytical instrument

(AI)

Master data flow

Functional integration

Trusted data

Application

Device

Sample prep(SP)

Addressable analytical instrument

(AAI)

Robotic instrument

(RI)


SCADA100ms

Advanced process control

(APC)1

sec

Process control

MetalAccounting

labProcess

control lab


The nature of the installed base often requires a combination of centralized and decentralized deployment of architectureThe nature of the installed base often requires a combination of centralized and decentralized deployment of architecture


systems


PrDB

EDW

OPM MA ERPMF

MPXMPX DBRefineries: PMR

MPXMPX DBRefineries: BRPM

MPXMPX DBSmelters: Polokwane

MPXMPX DBSmelters: Waterval & Mortimer

MPXMPX DBConcentrators: PPL section

MPXMPX DBConcentrators: Lebowa section

MPXMPX DBConcentrators: Rustenburg section

MPXMPX DBConcentrators: BRPM section

MPXMPX DBConcentrators: Union section

Field devices

Control systems


Supervisory systems

MPXMPX DB

PLC ObjectDB

RLC

WBLIMSPH Lab results

RICIN

SS

AFD WBAI

SP

AAI RI

RLC

SCADA APC

Concentrators: Amandelbult section

Reg

iona

l lev

el, d

eplo

yed

as c

lust

ers

Plan

t lev

elC

entr

aliz

ed

Its im

plem

enta

tion

of a

n “M

ES”

ofte

n fo

rms

the

mai

n th

rust

of a

n In

dust

rial I

T pr

ogra

m

Spec

ific

upgr

ades

to p

lant

hi

stor

ians

may

be

requ

ired

Func

tiona

lly

may

exi

st, b

ut

need

s da

ta to

be

sus

tain

able







5. Key MES requirements for the minerals processing industry


The Industrial IT architecture is driven by the main functions, input and output data flows in each layer of the architectureThe Industrial IT architecture is driven by the main functions, input and output data flows in each layer of the architecture

1. To execute control instructions received from the control system layerField devices

1. Controls unit processes, typically a selection of machine controllers and addressable field equipment.2. Collects and reports these values to a PLC supervisory layer (e.g. WINCC), or an APC (advanced process

control) system.

Control systems

1. Real time supervisory executable control logic to ensure reliable and optimal plant operation. 2. Continually calculates and provides set points for the control system layer. 3. Provides MES with granular detail on reasons for plant failure, based on root cause analysis.4. Deploy “soft sensors” where a physical measurements are either impossible or not yet deployed.5. Execute in accordance with recipe (either internally selected, or instructed by MES)6. Unit process level blending within the smelters, based on feed stock on hand and projected from MES level.

Supervisory systems

1. Provides a logical framework, based on linked objects, within which the execution of process steps within a plant are managed. Tracks the execution of these process steps.

2. Performs continuous metal balancing across key plant unit operations and the plant as a whole, as well as across the entire metallurgical complex.

3. Provides material tracking and genealogy functions, and assigns material properties to tractable batches(either physical or logical, e.g. time-based) at plant level.

4. Provides a maintenance management and execution function at plant level, based on detailed information about fault root cause analysis.

5. Accepts forecasts from metal forecasting as targets for a period and optimizes and controls plant performance to this target, considering local plant conditions.

6. Receives logistics scheduling instructions from a scheduling system and ensures its execution at plant level. Manage macro blending from concentrators to the smelters.

7. It provides operational reporting and analysis functions to plant personnel, to support in-the-moment decision making requiring access to detailed plant level performance information.

8. Support event-based integration to other systems (such as SAP, LIMS, WB, logistics scheduling)9. Enables Enterprise wide Energy Management

Manufacturing execution systems

Main functionsArchitecture layer


Field devices

Control systems



systems

Supervisory systems


Minerals processing

execution (MPX) functions

Minerals processing

execution (MPX) DB

Programmable logic controller

(PLC)Objects DB


Weighbridge(WB)

Laboratory information

management system (LIMS)

Plant Historian 10 sec Analytical results

Operational performance management

(OPM)

Metal accounting (MA)

Enterprise resource planning(ERP)

Metal forecasting (MF)

Processing database

Enterprise data warehouse

Robotic instrument

(RI)

Controller Interface

node

(CIN)

Soft sensor

(SS)

Addressable field device

(AFD)

Weigh bridge

(WB)

Analytical instrument

(AI)

Sample prep(SP)

Addressable analytical instrument

(AAI)

Robotic instrument

(RI)


SCADA100ms

Advanced process control

(APC)1

sec

Potential future:Group sample

and assay results management (GSARMS)

Integrated architectures requires different stakeholder groups with traditionally disparate agendas to work togetherIntegrated architectures requires different stakeholder groups with traditionally disparate agendas to work together

Process Control

Automation vendors

IT Mega vendors

CommercialIT

• Rational design is a prerequisite for success, but does not guarantee it

• Team dynamics and cultural design are vital

• Relationships need to be nurtured

Enterprise Architecture considerations governing the...

Documents

Transcript of Enterprise Architecture considerations governing the...