the integrated environmental monitoring (iem) project€¦ · classification: restricted iem...
TRANSCRIPT
IEM project Classification: Restricted
The Integrated Environmental Monitoring (IEM) project
Approach for development & implementation of a semantic model for the environmental domain
IEM project Classification: Restricted
Agenda
• Introduction Kaare Finbak, Kongsberg
• Ontology construction in IEM Morten Strand, DNV
• Semantic Challenges in Integrated Akram Bou-Ghannam & Environmental Monitoring Jone Aarre, IBM Approach to implementation of model
• Summary Kaare Finbak, Kongsberg
IEM project Classification: Restricted
The Integrated Environmental Monitoring (IEM)
project
Statoil R&D in Nov 2011 awarded Kongsberg, DNV & IBM a three year contract for joint development of a solution for Integrated Environmental Monitoring (IEM)
New concept
Proactive monitoring of environmental status
Part of day to day operations
IEM project Classification: Restricted
Main components of solution to be developed
Semantic Model for the IEM domain
5. Environmental data analysis
4. Information work spaces,
collaboration and work practices
3. Information access, interfaces,
integrity and security
2. Communications infrastructure
1. Sensor technology
IEM project Classification: Restricted
Depiction of development process
Scenarios Use cases
Work processes, methods & prototypes
Demonstra-tions
Life cycle focus
• Open area
• Seismic & drilling
• Installation
• Production
• Decommissioning
and after
Two first use cases
• Monitoring of drill
cutting discharges
• Production leak
detection
Prototypes
• Technologies to be
provided by IBM,
DNV and
Kongsberg
Structured evaluations
• Improvements to work
processes, methods &
prototypes
Dec 2011 Dec 2014
IEM project Classification: Restricted
Why a semantic model
• Need to present data from multiple sources in context
• Need to access data across domains, facilities, companies and institutions
• Want to have the opportunity to infer logical facts about instances
• Want to share environmental data with other companies, institutions and the public
IEM project Classification: Restricted
Goals of the semantic part of the IEM project
• Develop a semantic model for the environmental area
• Implement it in the solution that is to be developed, i.e., in the integration layer provided by IBM
• Demonstrate its ability to integrate data from various data sources and present meaningful data to applications
• Provide basis for future decisions on how to
• make model and environmental data available for others
• take the model forward to an approved standard
IEM project Classification: Restricted
Approaches discussed
Implementation of model
Development of model • Use-case driven ontology development
• Modular development
• Controlled use of external resources
• Apply appropriate Industry and W3C
standards as supported to instantiate
model
• Link to data to access authoritative data
• Map linked to data using model aware
approach
Access to model & data • Recommend how to take the model to an
approved standard
• Recommend how to make environmental
data available to others
To be presented
by DNV
To be presented
by IBM
IEM project Classification: Restricted
Development of the IEM ontology
Morten Rørvik Strand
Det Norske Veritas
IEM project Classification: Restricted
Motivation
The ontology will form the basis for the IEM integration model, and as such it must be robust enough to
• be the common nomenclature for all IEM activities and
• span multiple disciplines
• handle that many people will be involved in its development.
To do that, the IEM project must :
• Ensure that consistency is preserved.
• Tie-in to already existing and relevant domain libraries
• Be available and accessible
• Use the right open standards (Correct standards, OWL2)
IEM project Classification: Restricted
Content
• Use-case driven ontology development
• Tight cooperation between domain experts and ontology experts
• Frequent coordination with data sources and use cases
• Modular development
• Don’t let unrelated domains interfere with each other
• Ensure that common concepts are centrally defined
• Controlled use of external resources
• Be aware of implications of using external resources.
IEM project Classification: Restricted
schemas
Identify domains and consolidate information
Ontology Experts
Domain Experts
Nomenclature
knowledge
patterns raw data
Ontology
IEM project Classification: Restricted
Identify domains and consolidate information
Ontology Experts
Domain Experts
Ontology
Domain Data
IEM project Classification: Restricted
Modular ontologies: Importing
A
B
If a resource R is in A and B, we may say
Specific
ity • R is a subclass of X
• X is a subclass of R
• X is a subclass of R
• R is a subclass of X
IEM project Classification: Restricted
Selecting an Upper Ontology
Upper Ontology
Domain ontology A Domain ontology B
IEM project Classification: Restricted
The Ontology Stack: Partitioning of the Ontology
Upper Ontology
Core Ontology
Domain ontology A Domain ontology B
IEM project Classification: Restricted
External resources: Linking Data
JORD RDL
IEM Ontology Stack
Upper Ontology
Some ontology
Other public data source
NPD Fact Sheets
IEM project Classification: Restricted
External resources: Linked Data and reasoning
A
B
C – External
weak strong
weak
Reasoning along strong links
No reasoning along weak links
IEM project Classification: Restricted
The Whole Ontology Stack
etc
Upper Ontology
NPD
PCA RDL
Core
Domain 1
Domain 2
Domain 4 Domain 3
IEM project Classification: Restricted
Semantic Challenges in Integrated Environmental
Monitoring – Approach to implementation of model
IBM Industry Solutions
Akram Bou-Ghannam, Jone Aarre,
Semantic Days Conference, Sola
10 May 2012
IEM project Classification: Restricted
Harness the power of information through Integrated
Environmental Monitoring
SensorsSensors
Platform
RT
Platform
RTPlatform
RT
Platform
RT
ExternalExternal
Statoil
System
Statoil
System
Statoil
System
Statoil
System
Enterprise Service BusEnterprise Service Bus
IEMIEM
RTRTPlatform
RT
PlatformPlatform
RTRT
Statoil
System
s
Statoil
System
s
IEM KBIEM KB GISGIS
External
www
External
www
GODIGODI
Manage Model
Security
QualityAdapters Rule Engine
Access
Service Data Management
RSMRSM
AnalyticsAnalytics•Environmental
expertise
•Risk expertise
•Models
•Tools and
technology
Data Communication InfrastructureData Communication Infrastructure•Wireless network
•Cabled
•Power system
Workspace, collaboration and work practicesWorkspace, collaboration and work practices•Operation with RT monitoring
•Historical view of environmental data
Data Access and managementData Access and management
SensorsSensors
Platform
RT
Platform
RTPlatform
RT
Platform
RT
ExternalExternal
Statoil
System
Statoil
System
Statoil
System
Statoil
System
Enterprise Service BusEnterprise Service Bus
IEMIEM
RTRTPlatform
RT
PlatformPlatform
RTRT
Statoil
System
s
Statoil
System
s
IEM KBIEM KB GISGIS
External
www
External
www
GODIGODI
Manage Model
Security
QualityAdapters Rule Engine
Access
Service Data Management
RSMRSM
AnalyticsAnalytics•Environmental
expertise
•Risk expertise
•Models
•Tools and
technology
Data Communication InfrastructureData Communication Infrastructure•Wireless network
•Cabled
•Power system
Workspace, collaboration and work practicesWorkspace, collaboration and work practices•Operation with RT monitoring
•Historical view of environmental data
Data Access and managementData Access and management
• Operational Business Intelligence
• Standards-based Information Model [Reference Semantic Model]
– Provides much needed context for integration, interoperability, & Analytics
• Predictive Analytics
• Improve Operational Decision Making
• Anticipate operational risk Requirements, and execute in advance of critical points
– Monitor Assess Predict Optimize
Key enablers
IEM project Classification: Restricted
Environmental Monitoring Information Requirements
• Condition Monitoring requirements:
– Support operations, condition monitoring, measurements, data collection, diagnostics, early warning detection, equipment definitions, control, processes, interoperability and life-cycle events of a monitored entity
• Integration requirements
– Semantic Consistency
• Provide a semantically consistent view of IEM data across the extended enterprise
• Minimize uncertainty – maximize context
– Flexibility and ease of cross domain integration
• Accelerated integration using Linked Data and Semantic consistence as required by implementation platform
• Reduce costs of accommodating changes
• Flexibility to easily & quickly reorganize data to accommodate new business needs
• Flexibility of perspective in presenting IEM data
IEM project Classification: Restricted
IBM’s Information Integration Core (IIC) – A model aware
integration framework for managing Data Access in IEM
SensorsSensors
Platform
RT
Platform
RTPlatform
RT
Platform
RT
ExternalExternal
Statoil
System
Statoil
System
Statoil
System
Statoil
System
Enterprise Service BusEnterprise Service Bus
IEMIEM
RTRTPlatform
RT
PlatformPlatform
RTRT
Statoil
System
s
Statoil
System
s
IEM KBIEM KB GISGIS
External
www
External
www
GODIGODI
Manage Model
Security
QualityAdapters Rule Engine
Access
Service Data Management
RSMRSM
AnalyticsAnalytics•Environmental
expertise
•Risk expertise
•Models
•Tools and
technology
Data Communication InfrastructureData Communication Infrastructure•Wireless network
•Cabled
•Power system
Workspace, collaboration and work practicesWorkspace, collaboration and work practices•Operation with RT monitoring
•Historical view of environmental data
Data Access and managementData Access and management
SensorsSensors
Platform
RT
Platform
RTPlatform
RT
Platform
RT
ExternalExternal
Statoil
System
Statoil
System
Statoil
System
Statoil
System
Enterprise Service BusEnterprise Service Bus
IEMIEM
RTRTPlatform
RT
PlatformPlatform
RTRT
Statoil
System
s
Statoil
System
s
IEM KBIEM KB GISGIS
External
www
External
www
GODIGODI
Manage Model
Security
QualityAdapters Rule Engine
Access
Service Data Management
RSMRSM
AnalyticsAnalytics•Environmental
expertise
•Risk expertise
•Models
•Tools and
technology
Data Communication InfrastructureData Communication Infrastructure•Wireless network
•Cabled
•Power system
Workspace, collaboration and work practicesWorkspace, collaboration and work practices•Operation with RT monitoring
•Historical view of environmental data
Data Access and managementData Access and management
• Start with RSM as the meta model for IEM
– Hypothesis: IEM model requirements are very similar to the requirements for Condition Monitoring and Diagnostics (CM&D) of engineering assets
• Identify needed RSM extensions and link other models when & where needed
• IIC – a model aware integration platform
• Reference Semantic Model (RSM) – a key component
• RSM is built by linking multiple industry standards to form an enterprise wide Ontology
• RSM includes a Condition Monitoring ontology
Approach
IEM project Classification: Restricted
Machine Condition Assessment Process is the same as the
Environmental Condition Assessment Process
Monitor
Determine
Condition /
State
Assess
Health
Assess
Future
Health
Provide
Actionable
Information
Sensors Condition Monitoring
Systems Predictive O&M Solutions / Systems
What variability is preventing the asset from delivering the
performance it was meant to have?
DA DM SD HA PA AG
Is the asset operating as designed?
Diagnostics
What is the estimated time to failure based on projected
usage and historical data?
Maintenance
/ Operations
Processes
EAM, ERP
Prognostics
What are the recommended maintenance and
operational changes?
Advisories
Identify, predict, & manage asset operational risk:
1) Asset Health; 2) Environmental Health & Safety;
3) Process Safety; 4) Quality
IEM project Classification: Restricted
Extending the RSM to meet the Integrated Environmental
Monitoring requirements – A proposal
S 95 + S 88 S 95 + S 88
ISO 15926 ISO 15926
UN/CEFACT UN/CEFACT
CIM CIM
OPC OPC
MIMOSA MIMOSA
SI Base Units
SI Base Units
Reference Semantic Model
S 95 + S 88 RSM
ISO 15926 ISO 15926
UN/CEFACT ISA 95/88
CIM OPC
OPC UNCEFACT
MIMOSA Open O&M /
MIMOSA
SI Base Units
SI Base Units
1
Semantic Sensor
Network
2 • RSM – built by linking
multiple industry standards to form an enterprise wide ontology
• Condition Monitoring Ontology in RSM created from Open O&M / MIMOSA standards
• Link RSM with Semantic Sensor Network (SSN) ontologies
1
2
Others?
IEM project Classification: Restricted
Modeling Environmental Monitoring using RSM – Example
Hyme Field (Site)
H3 Coral Reef Area (Segment)
*
Temp. Transducer (Transducer)
Water Temperature (MeasurementLocation)
Measure
05 (Value)
Celsius (Value Type)
Water Temp (Measurement)
Registers
has
has
Lander (MeasurementSource)
Data: Measure
(Enterprise)
Registers Site
(AttributableEntity)
*
Segment (MonitoredEntity)
Registers
*
Asset (MonitoredEntity)
Registers
*
MeasurementLocation (AttributableEntity)
has
* Measurement
(HierarchicalEntity)
has
Transducer (Asset)
MeasurementSource (Asset)
0..1
0..1
Measurement Attributes:
-- Data: Is the actual measurement value and
could be of any Type specified in the
ValueContent enumeration
-- Start, Stop, & other time stamp attributes
-- Confidence, etc.
class 01 - Core Types
«XSDchoice»
ValueContent
+ BinaryData :BinaryData
+ BinaryObject :BinaryObject
+ Boolean :Boolean
+ CDF :CDF
+ Coordinate :Coordinate
+ HDF5 :HDF5
+ MathML3 :MathML3
+ Measure :Measure
+ MonetaryAmount :MonetaryAmount
+ Number :cct:NumericType
+ OCL :string
+ Percentage :Percentage
+ Probabil ity :Probabil ity
+ SVG :SVG
+ Text :cct:TextType
+ TextWithUnit :TextWithUnit
+ UTCDateTime :UTCDateTime
+ UUID :UUID
+ XHTML :XHTML
+ XMLAny :XMLAny
Enterprise (AttributableEntity)
*
Registers
*
Registers
IEM project Classification: Restricted
Summary
• Presented an approach for modeling the Real-time Environmental Monitoring domain
• Strategy: Build upon current (existing) RSM and extend it to satisfy the Environmental Monitoring domain.
• Rationale: Even though the RSM is build on standard ontologies from the Operations & Maintenance (O&M) domain - we feel it is generic and flexible enough to be extended and instantiated for IEM concepts. Specifically, the Condition Monitoring, Measurements, and data collection concepts.
• To Do:
– 1) Look into integrating the Semantic Sensor Network (SSN) standard ontology. Identify what parts of the SSN are needed and how to make the SSN and RSM ontologies consistent.
– 2) Identify needed RSM enhancement [Support for the updated UNCEFACT ontology, and additional needed Open O&M MIMOSA-CCOM concepts].