Title: Service design challenges in a safety critical time constrained domain
1Service design challenges in a safety critical
time constrained domain
- Einar Landre, StatoilHydro
- einla_at_statoilhydro.com
2Time and real world problems
Physical systems change with time
Control system must respondwithin defined time
envelope
Failure to respond within envelope lead to more
serious problems
Time an inherent property of thehandling process
Challenge Time is abstracted away
Oil well watering-out profile.
3Production control challenges
- Identified challenges
- Growing data volumes (information overload)
- No shared situation awareness
- No mutually agreed timeframes for action
4Conceptual Model
5Vision
- Situation characterized by
- Automated work processes, no information
overload - Shared situation awareness, humans focus on
important issues - System operates within defined time envelopes
6From processes to layers
7Layers Concerns
8ERP
- Production forecasts, financial control and
budgeted improvements - SAP, Schlumberger, Landmark, et al
- Highly manual, Aggregated Numbers.
- Month-end reporting
- Daily weekly reporting
9Automation
- Factory specific (40 sites)
- Process control
- Safety critical (gas detection, etc)
- Central Control Room
- ABB, Honeywell, Siemens, et al
- Produces time-series (tags) for others to consume
10Operations
- Distributed / Networked
- Offshore (Central Control Room)
- Onshore (Collaboration Room)
- Support Centres / External Vendors
- Highly Manual
- Information overload
- Time envelope violations
- Safety criticality
- Automation layer requirements into the
back-office
11Where are we?
New Apps
REST Data Services
OPC Data Services
SOAP Data Services
ESB
Reference Semantic Model (Virtualisation)
Historians Time-series
EC PI Plan
SAP Notifications
12Future
User Interaction Services
Networked Service Platform
Business Services
Business Services
Data Services
Data Services
Data Services
Reference Asset Model (Virtualisation)
Historians
SAP
EC
Process Control
13The challenges
- How to design networked services where time and
safety constraints are inherent? - How to model time in the context of networked
services? - How to address safety security in such
environment?
14References
- SPE 112078 Improving production by use of
autonomous systems, Intelligent Energy 2008 - SPE 103554-PP Software Agents An emergent
software technology that enables us to build more
dynamic, adaptable and robust systems, ATC 2006
15Our physical world
16System Perspectives
17A process engineers view