This Disaster Proves the Importance of Configuration Management

Configuration Management (CM) ensures an enterprise that all its products, processes, facilities, services, physical assets and IT systems are what they are intended to be and that they are properly optimized for their intended use. CM primarily tracks data records, documents and the links between files and data records; it also executes changes and modifications against those data items and links.

Like so many complex things, CM is better defined by what it does rather than what it is, as the following summaries show. So, this Deep Dive into the sixth of the 12 critical trends and enablers of digital transformation leans heavily on explanations.

The industry standard for Configuration Management, ANSI/EIA-649C, states that CM applies resources, processes and tools to establish and maintain consistency between the product requirements, the product itself and associated product-configuration information. “Product” in CM can be virtually anything, and ANSI/EIA-649C’s list is comprehensive: documents, facilities, firmware (embedded software), hardware, software, tools, materials, processes, services and systems. For specifics, read the SAE standard here.

Industry practitioners expand on this explanation.  

The Virginia organization CMPIC, formally known as the Configuration Management Process Improvement Center, notes that CM assures that products, facilities, IT systems, services, processes and more are what they are intended to be. CMPIC adds that CM assures that changes are properly evaluated, authorized and implemented while all information defining and managing configurations (or their data) is current, accurate, structured for all users’ needs and readily available to all who need to know.

The Institute for Process Excellence (IpX), is the purveyor of the CM2 methodology (a standard for Enterprise Configuration Management) and summarizes CM as managing products, facilities and processes by “managing their requirements.” IpX sees this as including conformance and relationships as well as changes, and that these are auditable, repeatable and verifiable. IpX observes that this is “keeping track of what you design, develop, deliver, sell and support” — a commendable and appropriate lifecycle view of CM.

CIMdata emphasizes that effective CM supports the reuse of standards and best practices while assuring that all requirements remain clear, concise and valid; and that any communication of relevant information is proactive and effective.

Given the complexity of CM, this level of detail is not overkill. Overkill is the number of terms entangled with CM; I have identified more than 30. Among them are change control, variant views, CI, effectivity and configurators.

Ultimately, CM is a methodology that seeks to ensure effective and efficient lifecycle management. CIMdata believes CM is vital for enabling good lifecycle-management environments; without CM, the management of lifecycles would be out of control. CM is also foundational to the validity of data and information in model-based systems engineering (MBSE), which in turn is a vital connection between CM and Product Lifecycle Management (PLM).

In everyday use, CM identifies and tracks basic problems that arise in:

  • Giving customers what they want
  • Creating new and derivative products
  • Making two products exactly alike
  • Managing changes to products
  • Pinpointing which part to replace
  • Fixing a broken product
  • Ensuring product safety
  • Managing bills of material (BOMs)

Within the CM processes themselves, many disruptions can be addressed:

  • Forced, rushed or unauthorized changes
  • Incomplete tracking of changes
  • Baselines missing or not updated
  • Overlooking/ignoring some users
  • Unknown first-time change impacts
  • Incomplete evaluations
  • Weak traceability of changes
  • Tedious change processes
  • Metrics not kept or not reported
  • Personnel using wrong information       

Deepwater Horizon Proves The ‘Why’ of Configuration Management

When effective CM is absent, negative consequences can be enormous. A case in point is the April 2010 explosion, fire and subsequent sinking of the Deepwater Horizon, a 33,000-ton offshore drilling platform in the Gulf of Mexico. The disaster was caused when the blowout preventer failed.

BP Oil, the platform’s operator, was held responsible for 11 deaths, although criminal charges were eventually dropped, as widely reported. In the three months it took to finally cap the well, millions of barrels of crude oil poured into Gulf waters.  As of a January 2018 Reuters article, according to BP Oil, cleanups on the coasts of five states, ongoing environmental damage, fines and pending claims would exceed $65 billion.  This sum includes a July 2015 settlement of state and federal claims totaling $19 billion.

White House investigations uncovered many failures that could have been greatly minimized or even eliminated had Deepwater Horizon’s owner, Transocean Ltd., provided BP with up-to-date information about the blowout preventer. BP engineers could not activate underwater safety equipment because it had been altered and BP's drawings did not match the device’s actual configuration. BP wasted hours figuring this out and days trying to stop the underwater gusher. Transocean said it did not change the blowout preventer drawings as it should have; moreover, the actual parts did not work, and the leak had to be plugged in other ways.

In retrospect, Deepwater Horizon points to questions that should be asked by any organization after significant changes in configuration. Does the change impact or affect:

  • Regulatory compliance?
  • Security?
  • Design, development or testing?
  • Product architectures?
  • Form, fit or function?
  • Hardware interchangeability, sustainability or replaceability?
  • Lineups of suppliers?
  • Critical or controlled items?
  • Performance or reliability?
  • Internal or external interfaces?
  • Work, product scope, delivery or schedule?
  • Equipment compatibilities?

And finally, does the change offer significant ROI?

With questions like these in mind, businesses invest in CM to gain more effective oversight of product development and production, sales and marketing, and support and maintenance. Amid the growing use of global virtual teams, businesses use CM to control access and ensure the sharing of accurate and current information.

Among the gains are better analyses of change impacts; better ability to create, find and use standardized parts; the reduction of part complexity; dealing with multiplying product variations and ever larger and more complicated products, systems and assets; mass customization and demands for personalization; greater design reuse and more beneficial use of Agile design as well as manufacturing and LEAN initiatives.

Additionally, CM helps minimize disruptive downstream changes and surprises. That translates into significant cost savings, speedier schedules, less production downtime and less scrap and rework. By helping organizations avoid costly errors caused by ad-hoc, erratic and unforeseen changes, CM makes product-lifecycle change processes proactive instead of reactive.

Using CM to maintain products with long lifecycles can, for example, stretch the life of airframes to 30 and even 50 years. Useful service lives of some other complex systems, e.g., aircraft carriers and submarines, can be supported even longer. In aerospace and defense, CM can also ease regulatory compliance. CM has been shown to enhance safety, increase data integrity and minimize product liability claims by preserving evidence of due care and diligence.

Within the organization, CM enhances productivity and efficiency by facilitating access to and understanding of common data, simplifying data synchronizations, easing the standardization/use of technology and accelerating time-to-market. My clients report big CM gains in measurable performance parameters, decision-making, production repeatability and quality assurance—while avoiding guesswork and trial and error. All these gains are part of digital transformation.

This conceptual view of CM shows the major considerations that must be addressed in the form of logical structures for an effective implementation. (Image courtesy of CIMdata.)

The ‘How’ of CM: Job and Authority Considerations

The manifold benefits of Configuration Management begin with managing and tracking changes from both internal and external sources, beginning to end (i.e., for the entire lifecycle). This means multiple CM-related logical structures must be established in service, sales and business planning, as well as in engineering and manufacturing. The necessary resources will be extensive. Many people will have key roles, not just CM professionals; in fact, the entire organization will have to contribute if effective CM is to be implemented and maintained.

Many aspects of CM are tied to specific jobs and authority levels. In manufacturing companies, engineers and product designers may have the biggest influence. They will need definitions of products, revision specifics for their documentation and baselines—starting points for changing configurations. Organizations will also want their engineering change notices (ECNs) and engineering change releases (ECRs) to be properly linked with their overall CM policies and procedures, as well as assurances that effectivity processes and dates will be properly communicated.

Elsewhere in the extended enterprise, differing job responsibilities have to be addressed:

  • Managers will demand a configuration control board for change approvals.
  • Service personnel will focus on ordering parts and kits, finding current and as-maintained configurations and locating technical documentation.
  • Customers will focus on ordering and maintaining preferred product versions.
  • IT will be concerned with repositories and libraries, data compression, the definitions of product information and data access and movement.
  • Finance will need to incorporate enterprise resource planning (ERP), where BOMs are generated and parent/child relationships managed.  

CM implementations become even more diverse when the specific needs of businesses—military contractors, aerospace companies, industrial/discrete manufacturers, process companies and utilities—are considered. Each also may demand unique approaches to risk analysis within CM.

At various points in any CM implementation, knotty questions arise. I suggest that project managers be prepared to designate configuration managers (and explain their roles), provide training and education, set up conflict resolution, resolve conflicting performance metrics (speed versus quality, for instance) and tackle arbitrary limits on product data such as mark-ups and test results. I also recommended that CM objectives be reflected in performance reviews.

Conclusion: Implementation is Crucial

Usually in a Deep Dive, I reserve the conclusion for summarizing benefits but the many possible uses of Configuration Management work against that here. So, instead I will look into implementation steps that can ensure benefits.

The obvious start is to determine scope. This requires identifying all users of product definition information throughout the enterprise, and where their CM needs converge and diverge. Ultimately, all the CM organizations, common processes, business rules, data formats and security strictures must be identified — along with the configuration of every item that CM will be managing.  

As with any process implementation, the attitudes of key users will determine ultimate success.  Levels of access to information shared across the extended enterprise will have to be defined; some persuasion and even some job re-engineering may be required.                

I have a few caveats:

  • Don’t be too eager to make a change to the way CM is handled without careful study of its impacts.
  • Match implementation to the needs of the organization, and choose appropriate levels of complexity and granularity.
  • Handle with care. CM spans entire lifecycles and many diverse parts of the organization.
  • Be prepared to change rules but stick to principles.

Compared to the potentially huge wastes of time and resources that accompany problems in products, processes, services and other assets, CM’s return on investment is significant when enabled by PLM. CM plays a critical role in lifecycle management ensuring that all data is clear, concise and valid. As a result, CM is indispensable to digital transformation.

Finally, it is important to remember that no one disputes that change is constant, so we must keep abreast of all product data modifications and be ready to find and fix them whether or not they are harmful. Changes and modifications can never be anticipated with any certainty, but Configuration Management gives us the tools to keep up.