UML Model Led Data Service Automation with Sparx EA

 1                  Overview

The Utilities CIM provides a common, objective terminology for electrical asset management, grid control, network data collection, and electronic sensors, devices and communications, as well as electricity market scheduling and financial information. It is a UML 2.x model, hosted in Sparx Systems Enterprise Architect.

Over the past decade, UML has played a dominant role in the development of IT architecture.

Sparx Systems is a leading UML model technology vendor.  Version 8 is a comprehensive platform, with extensive interoperability capabilities.  Dynamic visual filtering of diagrams ensures that EA is an easy-to-use environment for large complex UML models such as the Utilities CIM, and the technology of choice for governance of complex technology projects.

It is now possible to build  ICT data integration systems, directly from UML model technology, using comprehensive models like the Utilities CIM to define the data elements and interfaces.  

According to the Object Management Group website,  “a platform-independent MOF-based model is the secret to software stability and ROI “ .  Platform Independent  Models developed in UML, can be transformed into Platform Specific Models  for specific development environments.  This is a very exciting capability, and over the past few years, Sparx EA has provided add-ons for Model Driven Generation capability (MDG) to its portfolio -  “MDG Technology files wrap up Profiles, Patterns,Templates, Reference Data and more into a convenient single file for distribution to users and clients” (MDA Overview – Sparx Systems).

One Sparx MDG plug-in is a Data Distribution Service (DDS 2.0) based on the OMG specification. What is exciting about DDS is that it facilitates real-time data services over the wire, reducing the need for complex, resource consuming messaging at the application and presentation layers.  

DDS operates on the concept of pub/sub messaging at a domain level, with a simple key and topic structure that can be thought of as both database fields and primary keys, or XML element and attributes. 

And this structure is fairly simple to instantiate from a UML industry domain model, such as the Utilities CIM.

Figure 1: UML Models can drive real time distributed data services

At the same time in parallel, middleware technology suppliers have developed reference architectures based on information exchange models, to take distributed data services, and provide business intelligence to browser based end users.

The best-of-breed middleware services use common industry and integration models as the basis for exchanging information across system boundaries.

By mapping these domains to a technology model that recognises the CIM, any reference architecture that has a model exchange capability at its core, can be deployed to synchronise electricity network operational data stores, with scheduling and market data in real-time.

 2                  Energy Exchange Model

The Utilities CIM is a rich terminology for the physical details of the Utilities industry. To make this logical physical model deployable in current best practice ICT technology, links from the CIM to the integration metadata are required to deploy a UML model in MDA style technology deployment.

Figure 2: Utilities CIM and Energy market integration metadata example

Information exchange is not only predicated on industry function and data, but also identification of workflow, devices, events and other contextual information, such as geospatial routes, fences and maps.

An overview Energy Exchange Model , illustrates the metadata required to deploy the Utilities CIM as real-time distributed data services.

Specific electrical network and market information can be connected from the CIM to  technical and business metadata. For example technical data such as network device ID and network address, enterprise data such as location information and job role, and data from external sources such as market and geospatial services, etc., can be contextually accessed in real-time  to add value and meaning to electricity supply, energy management, market and scheduling data. 

Real-time analytics for both operational and business intelligence can be facilitated, enabling coherence of SCADA, EMS and other network device data, to add geospatial overlays, such as INSPIRE, to add location, and interfaces with ERP and reporting systems, etc.  

To enable workflow to make use of both operational data and technology metadata, the electricity and technology integration domains have to be logically linked.  A metamodel can also link information across CIM domains, e.g. providing Network Outage and Asset Management links, by providing keys to cross domain topics.

To put the whole picture in context, the Utilities CIM with a  linked integration metamodel, can generate the platform specific model for the data distribution services using an MDG add-in for DDS2.0 from Sparx EA.

The same Utilities CIM plus metamodel can be used by Progress Software as a common model exchange for high speed complex event processing of electricity network data with market data.

By using the CIM and the metamodel as the basis for the exchange, business users can run workflows that trigger the events  that allow end users to view market demand data, in a geospatial context, in real time.

In summary, Smart Grid can be enabled by using a common model for network data collection and distribution,   and real-time geospatially aware data access from Energy Management Systems, business metadata, and corporate systems such as SAP and reporting.

Cost effective system delivery is ensured by virtue of using the same common logical model for both  data collected from electricity supply and management systems inputs, as for integration, operational and business intelligence outputs.  The deployment can be made to work across borders and regions, utilising geospatial context in real-time complex event processing.  

There is every reason to utilise the same models across all Transmission Network Service Providers in a cross-region, cross-border energy grid.  This would provide for the ability to share 'Smart Grid' electricity network information, readily and cost-effectively amongst consumers, generators, transmitters, distributors and the market.

 The OMG DDS specification is designed to allow UML led data services on distributed infrastructure, and this could be applied across all stakeholder TNSPs.  With core common UML models, and an infrastructure cloud with network layer security, wide area energy demand management services may be delivered simply and at relatively low cost.