slide 1 standards harmonization through semantics unification joe zhou xtensible solutions fran...
TRANSCRIPT
Slide 1
Standards Harmonization Through Standards Harmonization Through Semantics Unification Semantics Unification
Joe ZhouXtensible Solutions
Fran ClarkArpeggio Technology, LLC
Progress Software
Denver, CO November 18th, 2009
Slide 2
TopicsTopics
• Introduction – Presenters and company background
• Problem/Opportunity Statement – The evolution of Smart Grid will require a scalable strategy
and approach for standards development and adoption.
• Smart Grid Requires Smart Data – Common Semantic Model for Smart Grid (CSM-SG) will get
you there.
• Lessons Learned from Other Industries• Summary
Slide 3
Problem / OpportunityProblem / Opportunity
• Smart Grid initiatives present an unprecedented opportunity for the energy industry to serve its customers with a new generation of efficient and reliable products and services.
• Smart Grid implementations will require: – Capture intelligent integration of detailed data from a large
and heterogeneous group of sources– Informed cooperation amongst all the actors (providers,
distributors, customers, product manufacturers, 3rd part vendors, etc.)
– Ability to evolve rapidly as the technology, marketplace, etc. also changes.
– Standards-based “lingua franca “ to facilitate data exchange
Slide 4
Current Standards Current Standards
• Current Standards State to Support Smart Grid– NIST has identified a large number of existing standards needed for
Smart Grid (31 for initial list and 46 for future considerations). – Many of these standards are or contain data related specifications,
and large overlaps exist between them. – Majority of the existing standards require further enhancement to
support Smart Grid needs.– Existing standards address different levels of the interoperability and
cover different domains with large overlaps, which presents a huge challenge for harmonization and evolution: • Differing levels of maturity • Overlapping areas of focus• Different level of physical implementation (application vs. device)
Slide 5
New Approach NeededNew Approach Needed
• Standards must be harmonized in a way that – Leverages a shared semantics – Provides mechanisms to rapidly evolve (new versions of
standards, new standards altogether)
• Implementation of standards for data integration need to
– Reflect the specific needs and innovations of the implementing organizations
– Retain traceable lineage to official published versions of the standards
Slide 6
The Need for a Common Semantic The Need for a Common Semantic Model for Smart GridModel for Smart Grid
Source: Microsoft Smart Energy Reference Architecture
A Common Semantic Model for Smart Grid will emerge to unify the information definitions at the intersection of all Smart Grid domains.
Slide 7
Common Semantic Model for Smart Grid Common Semantic Model for Smart Grid (CSM-SG)(CSM-SG)
• Semantic Model presents a singular, non-redundant view of all the business concepts and relationships between those concepts
• Semantic Model is used to:• Map to concepts in existing concepts in existing
standards – Rationalize and Harmonize• Generate new standards to “fill in gaps”• Provide a stable core to support the generation of new
standards and new versions of existing standards• By organizations to create individuated Enterprise
Semantic Model (ESM)
Slide 8
Semantic Model-Driven ApproachSemantic Model-Driven Approach
Semantic Modeling
SemanticMapping
Large Scale Enterprise Integration
Focus is on Semantics
Meta-data
Enterprise Model
Data Services
Data Services
Exchange Model
Slide 9
Model-Driven Approach Model-Driven Approach End-to-EndEnd-to-End
Separation of Semantic/Logical Model from Physical Models needed to support integration
Physical Models to represent:– Published wire-format standards– Legacy, Vendor APIs – Organization specific interfaces based on ESM– Multiple versions of each
Exchange Model– Modeling constructs to represent mapping
between each Physical Model and the “hub” Semantic Model (CSM/SG or ESM)
– Unambiguous documentation of mapping rules
Process, tools and methodology to generate artifacts based on the Semantic Model
– WSDL / XSD– DDL for RDB– etc.
Data Services to implement integration
– Enforce / leverage Semantic Model
– “Makes it real” • models that are used for
implementations are much more likely to stay current and useful
• directly leverages investment in modeling
– Standards Bodies• Validation / certification
services– Implementing organizations
• Validation• Transformation• Semantic Mediation
Slide 10
Model-Driven Process – End To EndModel-Driven Process – End To End
Requirements Design Development Maintain
UM
LM
odels
• Process Models• Integration
Requirements• Semantics
• Canonical Models• Service Models
• Canonical Messages• Service Definitions
• Source UML model–Behavioral–Structural
ExchangeM
odel
• Interface Comparison
• Gap Analysis
• Canonical Data Services
• Non-canonical Models
• Data Service Runtime• Data Service Unit
Testing
• Source model for executable components
Dev.Env
• Communication, Collaboration
• Documentation• Architecture• Infrastructure
• Additional run-time components, transport, applications, etc.
• Repository for builds, environments, control
Slide 11
Solution Benefits
• Harmonization makes Smart Grid possible • Semantic Model allows for harmonization today (or
soon) and evolution tomorrow • Process provides governance and compliance • End to end model-driven approach
– Reduces effort required (design, implementation and maintenance)
– Speeds time to market– Optimally identifies and manages change
Slide 12
Complementary ProcessesComplementary Processes
Standards Process Enterprise Process
Industry Semantic
Model
Enterprise Semantic
Model
ExchangeModel
Data Services(Transformation,
Mediation)
Slide 13
Lessons Learned From Other Industries
• Lessons Learned From Other Industries – Finance (SWIFT/FIX)
• SWIFT and related standards – attempt to converge around ISO20022 standard
• SWIFT based integration implementation– Insurance (ACORD)
• ACORD Information Model standards mapping effort• ACORD based integration implementations
– Telecommunications (SID) • SID standards development and model dissemination• SID based integration implementations
Slide 14UNIFI_(ISO_20022)_v32
ObjectiveTo enable communication interoperability between financial institutions, their market infrastructures and their end-user communities
Major obstacleNumerous overlapping standardisation initiatives looking at XML financial messages:
MDDL, FIX, FinXML, VRXML, RIXML, XBRL, FpML, IFX, TWIST,
SWIFT, RosettaNet, OAGi, ACORD, CIDX, etc.
Lessons Learned: FinanceLessons Learned: FinanceUNIFI – ISO 20022UNIFI – ISO 20022
Proposed solutionA single standardisation approach (methodology, process, repository) to be used by all financial standards initiatives
UNIFI (ISO 20022)
ISO 20022 - UNIversal Financial Industry message scheme, the recipe:syntax neutral business modeling methodologysyntax specific design rulesindustry led development/registration process financial repository on www.iso20022.org reverse engineering approach to ease coexistence
Slide 15UNIFI_(ISO_20022)_v32
Growth adds exponential complexity and expense…
RosettaNet
OAGi
TWIST
Proprietary format
SWIFTIFX
EDIFACT Without common building blocks:• Point-to-point connection• Data is mapped directly from one application to another• Costly, unscalable and difficult to implement and maintain• Process, routing, rules logic needs to be coded to specific message types42 interfaces = n * (n-1)
Source: John Mersberg, IBM Corporation
UNIFI – ISO 20022UNIFI – ISO 20022
Slide 16UNIFI_(ISO_20022)_v32
Standardised implementation reduces cost, time to effect change and improves overall performance…
Canonical message model =• True process integration• Reduced brittleness, faster to respond to change• Shared message services – single/shared parser, message independent rules engine, etc.• Unified monitoring / audit trail
RosettaNet
TWIST
SWIFTIFX
EDIFACT
Canonical Message Model(i.e. ISO 20022)
UNIFI aims at long term convergence, while facilitating short term coexistence…
14 interfaces = n * 2
Source: John Mersberg, IBM Corporation
OAGi Proprietary format
UNIFI – ISO 20022UNIFI – ISO 20022
Slide 17UNIFI_(ISO_20022)_v32
All institutions have their own sets of data objects
ISO standardises common data objects…
Account
Order
Date
…and groups them into ‘syntax-neutral’ message models, which...
Order
DateDate
XML ISO 15022
… can be ‘transformed’ in message formats in the desired syntax
FIX
EDIFACT
UNIFI – ISO 20022UNIFI – ISO 20022
Slide 18
Summary / Next StepsSummary / Next Steps
• Smart Grid initiatives present an unprecedented opportunity that will require unprecedented interoperability.
• Interoperability is a journey not a destination, which requires:
– Harmonized Standards– Mechanism for Managing Standards Evolution – Common Semantic Model (Standards and Enterprise)– End-to-End Model-Driven Approach to Data Integration
Slide 19
Thank YouThank You
• Joe ZhouXtensible Solutions – [email protected]
• Fran ClarkProgress Software – [email protected] Technology, LLC – [email protected]