Integration of control, manufacturing and enterprise systems

Transcription

1 1 Integration of control, manufacturing and enterprise systems Chris Hamlin, SABIC UK Petrochemicals Nina Thornhill, Imperial College London Outline of talk 2 The broad picture ISA-95 Integration between PCS, MES and ERP General comments Experience from SABIC the new polyethylene plant on Teesside Discussion on future research directions 1

2 3 The Broad Picture The broad picture ISA-95 4 The computer-integrated manufacturing (CIM) pyramid of the 1980s has crumbled to make way for.. better models for manufacturing information technology in the 2000s. Keith Unger, InTech,, 03 October 2001 Scheduling Optimization Advanced process control Enterprise resource planning, ERP Manufacturing execution system, MES ISA-95 Enterprise Control System Integration Regulatory control Process control Control valves, sensors, hardware system, PCS 2

3 The broad picture ISA-95 5 ERP Demand Sales & Distribution Enterprise Performance Reporting Supply Chain Material Warehouse Cost Accounting Asset Human Resource Maintenance Business & Logistics MES Scheduling Process Optimization Advanced Control Performance Process Analysis Manufacturing Operations Asset Monitoring Execution Material History Quality PCS Real-Time Execution Real-Time SPC Process Control Real-Time Control Real-Time Monitoring Adapted from: Sham Afzalpurkar CEO, Performix Inc., 2006 ASUG Annual Conference 6 Integration between PCS, MES and ERP 3

4 Tyranny of NOW! Control systems have evolved to manage the present Advanced control uses historic data to make deterministic predictions of the future but only to decide what to do straight away (Real-time) optimisation is used to drive to a deterministic ideal for a instant in time Irrespective of the consequence over time Without any regard for inherent uncertainty 7 Scheduling is generally about managing deviations and opportunities Establishing the theoretical ideal is the least of the challenges Scheduling is inherently about managing risk and Optimization uncertainty Advanced The hierarchical translation of planning outcomes process control inevitably loses all information about the original Regulatory control intent Control valves, sensors, hardware Tyranny of Retrospection 8 Conventionally Fixed interval, retrospective, restricted content reporting Fixed intervals driven by financial reporting requirements Retrospective due to requirement to collate & reconcile information ion Content restricted by pre-defined format displaying all information available Leaders in our industry use technology to Monitor everything that might be important in real-time Displaying what matters, to those who need to know, when they need to know it, at all levels in the organisation Ensure planning, reporting and improvements are driven through a common, coherent information framework No value disappears between the cracks of inconsistency 4

5 Business Control Hierarchy Risk Coping with an uncertain World Prices & markets Supply and demand uncertainty Plant availability Agility Responding to fluctuations and deviations Alternate production and supply routes Spot opportunities Responsiveness & flexibility Enterprise resource planning, ERP 9 Performance Knowing what s s happening and what s s possible Identifying deviations (in production performance, supply/demand patterns etc) Validation of planning models and assumptions Driving continuous improvement Manufacturing execution system, MES Conformance Doing what s s wanted or expected Staying on spec (quality control) Minimising costs (constraint control) Pushing appropriate limits (real-time optimisation) Conservation Process control Keeping the process running system, PCS Basic regulatory control (closed-loop) loop) Conservation of mass and energy Competitive Advantage 10 Formula years ago Technology was primary source of competitive advantage Focus on automation Traction control Active suspension Automatic transmission Launch control Intelligent braking Fly-by by-wire 5

6 Requires Technological Advantage 11 Formula 1 today: Automation is an enabler No longer differentiates Technology remains a source of competitive advantage Emphasis now on real- time decision support No longer need to wait until the end of the race to understand what happened So what have we done on Teesside? Risk SAP ERP System Logistics Supplier MES Batch Genealogy System 12 LIMS order n-2 order n-1 Confirmed Quantity Quality Data Working production order n-2 Confirmed production Working production order n-1performance Batch characteristics Historian Working production order n to date order n Consumption Consumption to date Agility Working production order n+1 - Standard Tasks Recipe to task work flow Task Scheduling & Assignment Working task status handling Plant Info order n+1 Grade Quantity Grade to batch recipe mapping table Conformance Recipe Recipe Translation Quantity Working recipe Quantities Parameters Setpoints Profiles Times Limits Task Monitoring & Control Task y waiting Task y+1 waiting Task y+2 released Task y+3 completed Other Other order n+2 DCS Batch Control Logic Plant Control Working recipe Process Measurements Conservation 6

7 Learning & Challenges 13 Integration of different layers How to retain the core message, ensure coherence and prevent inconsistency Treatment of time Immediate and instantaneous at plant level Integral and multi-versioned at business level Fit with existing business models Comparing apples with apples (consistency vs accuracy) Incremental change across entire business of Information Data structures Visualisation Maintenance 17 END 7