By Greg Robinson, Xtensible Solutions
Most large electric utilities have had at least one project “try out” the industry-standard Common Information Model (CIM). A number have even based their large-scale enterprise application integration efforts on it. While these efforts are generally successful, the additional costs (real and perceived) often make people question whether the benefits are worth the trouble. The answer is a definitive: “It depends.”
Before a utility can make an informed decision regarding when and how to use the CIM, it needs to (1) have a clear understanding of the goals the utility intends to accomplish with CIM; and (2) plan accordingly with its eyes open to the actual state of and plans around CIM standardization. To the casual observer, CIM standard development efforts seem never-ending; they continue to progress and evolve. Therefore, numerous issues must be managed by end users and their service providers.
When people refer to the CIM, they usually mean the information model itself as well as all of the associated IEC 61968 and IEC 61970 standards that support its use (at www.iec.ch, refer to working groups 14 and 13, respectively, of Technical Committee 57). The CIM itself is defined in Unified Modeling Language (UML) notation and consists of classes and attributes for these classes, as well as the relationships among them. A key purpose of the CIM is to provide a common language to describe exactly what data is being exchanged among a utility’s business systems. For example, as opposed to using custom tags in XML messages, field names are based on class/attribute and association relationships defined in the CIM.
Interest in CIM use has recently surged as evidenced by the formation of and participation level in the CIM User Group (www.ucainternational.org). This growing interest is somewhat attributable to the evolution of systems integration technologies from Enterprise Application Integration (EAI) to Service Oriented Architecture (SOA) and Web Services, as well as the increasing needs from utility businesses to have better information management. As utilities continue to look for total business integration solutions, regardless of underlying technologies and standards, the CIM and related standards will likely play a critical role in transforming today’s point-to-point integration landscape to service-enabled enterprises. But how does one go about this?
Two Approaches for Using the CIM
There are at least two vastly different views on why and how a utility should use the CIM, resulting in a tug of war that can hinder productivity.
- Minimalist: Information management minimalists have a tactical, project-oriented focus and prefer to use the CIM on a project-by-project basis for specific system interfacesonly when it lowers project costs and/or risks. Many view themselves as protecting projects from overly burdensome corporate standards and constraints that would impede progress on “real work.”
- Strategic: Enterprise Information Management (EIM) champions have an overarching, long-term focus and see the CIM as a key component for IT to enable business transformation and performance optimization. They view information as a corporate asset and want to protect the enterprise from “cowboy” projects that create artifacts that are expensive to maintain and hard for other projects to use.
The advantage to the minimalist view is that utilities can achieve it with minimal IT master planning and little upfront investment. IT projects may continue operating in much the same manner as they always have, and everyone knows their roles. Projects and vendors have more autonomy and more control over their work. It is the natural default position, and since it is more comfortable for many parties, it is a common approach even though a utility’s management has not made a conscious decision on this matter.
On the other hand, some utilities opt for the second view because they are being compelled to create a more flexible business and are tired of their plans being hindered by complex, brittle IT systems. For example, some of these utilities are now planning their investments in such a way as to get more intelligence and value from the data that will be collected from “smart grid” devices, especially automated metering infrastructure (AMI). Using data intelligently wherever it is needed throughout the enterprise requires a new architecture and strategy. This leads these utilities to an integration strategy that is based not only on standards such as the CIM, but also EIM to give the standards a proper context.
Gartner has defined EIM as a commitment to:
- structure, secure and improve the accuracy and integrity of information assets;
- solve semantic inconsistencies across all boundaries;
- and support the technical, operational and business objectives within the organization’s enterprise architecture strategy.
EIM would ensure that a utility’s service providers each accomplish their individual functions in a manner that positively contributes to enterprise objectives. As depicted in Figure 1 (page 60), EIM is defined through a framework that encompasses five major components: vision and strategy, governance, core processes, organization, and infrastructure.
This second view (CIM in an EIM context) requires planning that is driven by business strategy and implemented in both a top-down and bottom-up fashion. As it represents significant change, many restraining forces can hinder its adoption. A utility must be prepared to proactively deal with both the driving forces and restraining forces when it embarks on the strategic view.
Driving Forces, Restraining Forces
Driving forces are those which help achieve a goal, in this case organizational adoption of the CIM in an EIM context:
- Consistent enterprise-wide data. Data is consistent and congruent irrespective of data sources. For example, there may be legacy and/or new work management systems for substations, others for distribution, others for plants, and so on. Additionally, different systems may be used among the service territories. If two utilities have merged, even more of the same types of systems will exist. The same type of information ought to be represented the same way for all work orders regardless of the system generating it.
- One version of the truth. Information about assets will exist simultaneously in many systems (work management, facilities management, design, financial, SCADA, network management, protection, etc.). This data should be consistent among these many systems and data warehouses so that operations are efficient, hazards are avoided, and so the data can be properly used for business intelligence and decision support.
- Access to data regardless of source. Data is made easy to discover and access through a consistent set of services irrespective of the data’s source.
- Business transformation agility. Implementing an EIM-based integration architecture improves IT flexibility, which serves to facilitate business change (rather than hinder it with a complex and brittle integration of systems).
- Reduced project implementation costs. While it takes time to learn pre-built industry information models, canonical message models, use cases, etc., doing so results in a net productivity gain. Projects can shift more time to the application at hand and less on determining how to integrate applications.
- Reduced maintenance costs. In terms of both time and money, it is expensive to support a different set of interfaces emanating from each application system. With an EIM in place, vendor-proprietary technologies and methods are hidden behind well-defined interfaces. This enables applications to be upgraded or replaced with significantly less effort and without the side-effects that commonly occur when traditional integration techniques are used.
- Reduced IT risks. The risk of missing integration requirements is significantly reduced as the CIM is used in multiple utilities in multiple domains with multiple application systems, with numerous parties feeding corrections and lessons learned back to the IEC for incorporation in future releases.
- Availability of external services. While standard interfaces have proved desirable within the enterprise, it is imperative for business to business integration. Using the same semantics across both minimizes development costs as well as the risk of errors in transformations spanning different business worlds.
- Scalable business process automation. Consistent data enables business process automation to be implemented independent of the number of applications and data stores involved. To do otherwise results in brittle and expensive solutions that will not scale to any significant size. An EIM-based semantic layer is required for scalable business process automation, monitoring and management.
- Scalable business activity monitoring. (same rationale as for ‘scalable business process automation’).
- Accurate reportingregulatory, KPIs. (same rationale as for ‘scalable business process automation’).
- Mergers and acquisitions. An EIM-based integration framework provides a neutral vehicle for integrating disparate applications from different companies. Decisions to replace individual applications can be deferred until there is a business need, rather than doing so merely for merger purposes.
Some common restraining forces that will tend to hinder adoption of CIM in an EIM context include
- Lack of stable integration standards. As previously mentioned, the CIM and other relevant standards continue to evolve, thereby requiring significant effort on version management.
- Vendor’s way = lower project costs. Vendors will want to use their methods, tools and data models for interfacing with other systems. In the absence of utility required methods, tools and governance, vendors will make the case to the utility’s project team that their way offers the lowest cost and risk.
- Vendors pushing for customer lock-in. Many vendors employ strategies to achieve “customer lock-in” through use of proprietary data models, methods and tools. While they may agree to implement standard interfaces, these vendors will often find ways to influence the customer to use proprietary interfaces.
- Consultants pushing to be “thought leaders.” If a consultant is not an EIM or CIM expert, he or she knows his/her value will be diminished on projects. Faced with this threat, some will take a position that, “while these are good concepts, they are not practical to implement. You’re much better off using our field-proven approach.”
- “Hours-sold” revenue driving system integrators. More efficient methods and the ability to swap in other companies’ resources would seem to be at odds with a well-entrenched system integrator that receives revenue based on hours sold to the utility. Some system integrators are embracing the CIM because it provides an opportunity to build a competitive advantage into their service offering, making them more attractive for future opportunities. Others will view it as a threat to their revenue stream and will want to effectively kill substantial use of it.
- Internal experts want to remain experts. Some experts of legacy systems or technologies that are subject to being replaced by the new systems/approaches can be most innovative at coming up with reasons why the new approach will fail. This is yet another form of “lock-in.” Without staff capable of separating legitimate concerns from emotional reactions, projects can make poor decisions resulting in substantial additional costs and schedule slippages.
- Project managers striving for control. Of course, it is desirable for a project manager to mitigate risks by ensuring he or she has good controls on the project. So, unless a manager can rely on an effective corporate integration framework with the necessary standards and governance, the manager will naturally push much responsibility onto vendors. Unfortunately, this drives them right where the vendors want them to be: “vendor’s way = lower project costs” and “vendors pushing for customer lock-in” discussed above.
- Inertia. Many people are comfortable with the way things are and would prefer not to change. A project challenge is to get these people to see how the change will be smoothly implemented and that the new world will be a better place for them.
- Our situation is unique; standards hinder us. This is a common belief that needs to be openly worked through. Experience suggests that in almost all cases, industry standards can support these special needs with only a modest amount of extension. These claims usually dissipate when someone with sufficient expertise about the standards is part of the team.
If the CIM and associated standards were finished and stable, a utility might be able to start off with the minimalist view and later switch to the strategic view. Unfortunately since the CIM glass is only half full, the amount of “some assembly required” is still significant. There is a considerable amount of wiggle room in how the CIM and associated standards can be implemented; and with so many parties involved, deliverables from disparate projects will not properly fit together. Therefore, significant rework will almost certainly be required before CIM-based artifacts from disparate projects can be aligned to contribute to the enterprise information management strategy. The best way to deal with this is to proactively plan EIM around the CIM and other relevant information standards.
Which Way to Go?
A utility needs to have a clear understanding of the goals it intends to accomplish with the CIM before it decides when and how to go about using it. To do otherwise would result in a poor price/performance ratio for its integration investments. The safest bet is to use a “minimalist” approach until the utility has truly come to grips with its information management objectives. This happens when a utility’s management gets serious about managing data so that it becomes a business enabler rather than an inhibiter. To implement this strategic approach requires business units and IT to look at data from the perspective of Enterprise Information Management (EIM).
Using CIM in the context of EIM would ensure that a utility’s service providers would each accomplish their deliverables in a manner that positively contributes to enterprise information management objectives. But because data integration involves so many internal and external organizations, there will be plenty of opportunities to derail meaningful progress. So, when the time comes to implement this strategic approach, the utility management will need to convey to various stakeholders that it is serious about information management by having solid, but not overly burdensome, governance in place.
As the CIM will play a key role, EIM planning should only be done with working CIM knowledge on-hand, because at the core of EIM is the management of industry standards-based semantics. So the utility should have team members that have significant experience in: (1) applying the CIM in the context of EIM and (2) successfully applying the CIM on utility projects that have gone into production. Since the CIM will continue to evolve, ideally these team members would have active ties with IEC working groups responsible for CIM development. Since information management is so strategic, mistakes will ultimately be costly and frustrating. So take the steps to be informed and ensure you have a good handle on how to suitably leverage the CIM.
Greg Robinson is a co-founder and President/CEO of Xtensible Solutions, which provides enterprise information management and integration solutions and services to the energy and utility industry. Greg is the international convener of IEC TC57 Working Group 14, which is extending the industry standard Common Information Model (CIM) for enterprise-wide messaging. He has a BSEE from Georgia Tech and a MBA from the Florida Institute of Technology.