A Tech-Management "Third Rail": Aerospace and Defense Group Seeks to Mitigate High Costs and Risks of Lost Data in PLM Obsolescence

It's always risky to drag metaphors from one realm into another, but sometimes the urge is irresistible. Take the matter of technology obsolescence, traditionally the Third Rail for information technology (IT) management.

Third rails have delivered electric power to subways for over a century, and they are dangerous. The usefulness of this particular metaphor in obsolescence of data management solutions and maintaining legacy solutions is the voltage in the third rails.

If a PLM solution becomes obsolete, the complexity in disconnecting and removing such legacy solutions can be as risky for the business as touching a third rail.

Obsolescence has always been a part of technology, as this image of an old Cray supercomputer reminds us. In a few decades, Moore’s Law has shrunk the computer hardware powering this Cray into electronic devices too small to be seen with the naked eye. Analysts may disagree on the metrics but corollaries of Moore’s Law apply to software as well. (Image copyright Microsoft Corp.)

Because so many data management applications in aerospace and defense companies are tightly linked, the choice to upgrade any one of them can raise dozens of issues—issues that busy managers have little time for. This is why such obsolescence decisions are often no decision: “If it's not broken, don't fix it.” Putting off the “day of reckoning” ultimately guarantees obsolescence and ensures process disconnects.

If solutions are designed and implemented in a sustainable way, the obsolescence risk can be minimized and may be even avoided. CIMdata is helping Product Lifecycle Management (PLM) solution users overcome the challenges of obsolescence—real and potential—in a myriad of lifecycle solutions, systems, and tools on which aerospace and defense (A&D) companies rely so heavily. This frees up IT managers to add value to the business.

Unfortunately, to some IT managers, many solution upgrades look like touching a Third Rail. The metaphor underscores that some IT managers see career risks lurking in dealing with obsolescence solutions and legacies. As to why these are still running, there is no simple answer.

Long-in-the-tooth tech has rarely been discussed in public forums. Typically, the most promising forums for obsolescence and sustainability are sponsored by solution providers who often see obsolete applications as uncomfortable topics. As a result, an entire industry—data translation—has grown up to deal with data in obsolete formats and applications.

An indication of the size of the obsolescence and sustainability challenges was in a PLM Road Map™ presentation by Jeff Plant of Boeing Commercial Airplanes. As Chief Process Architect and Director of Business Capabilities, he pointed out that Boeing's commercial airliner operations use 8,500 different computerized systems. Boeing wants to cut this to about a thousand, he said.

Research identifies the unacceptably high costs of solution upgrades and the accompanying risks of data loss. This research, carried out by the CIMdata A&D PLM Action Group, ranked the causes and notes relationships between them. Details of the program are spelled out in the Action Group's report “PLM Obsolescence Management, Phase 1 Research Report: Initial Model and Current State Assessment.”

The two major consequences of obsolescence are the Cost of Technology Refresh (the left two columns) and the Risk of Product Data Loss, the columns on the right. The left-hand stack in each column lists top factors that increase costs and risks. The right-hand stack shows the top factors that mitigate or reduce cost or risk. (Image courtesy of CIMdata A&D PLM Action Group.)

In this particular Action Group, PLM executives from five leading A&D companies are supporting research that define policy, process, and technology guidelines along with requirements to minimize the cost and risk of data loss associated with PLM technology obsolescence while keeping their enterprise PLM solutions refreshed and up to date. In addition to Boeing, members are Airbus, Gulfstream, Embraer, and Rolls-Royce. Membership is open to commercial aircraft original equipment manufacturers (OEMs), defense OEMs (aeronautics and space sectors only), and aircraft engine providers.

Phase 2 of the obsolescence research, currently underway, focuses on the relationships between causes and potential mitigating factors. Preliminary findings of Phase 2 include identification of benefits from using advanced PLM features within design teams, issues caused downstream by using advanced features, and how standards such as STEP support the business processes and mitigations.

Ultimately, this research will provide data to support direction statements that will be presented to PLM solution providers and standards bodies. The direction statements will provide clear communication from a single voice to help the solution providers and standards bodies address the PLM issues the A&D industry is facing.

From the A&D PLM Action Group's research so far, two takeaways have emerged:

As technology users, Group members must step up to their primary role in mitigating the costs and risks of ongoing technology refreshes.
As technology consumers, Group members recognize that solution providers will play significant supporting roles, but not until they have clear communication and strong financial incentives from customers.

To conclude, there is (as always) bad news and good news.

First the “bad”: PLM upgrades won't be getting significantly easier for a while. The basic reason, as the research shows, is that the issues are rarely clear-cut. Three examples:

Industry standards, many of which are supported only in part by solution providers and not widely used by industry anyway. Standards lag even widely deployed technology. Comprehensive data aging plans also fall short of their hype.
Long-term archiving, a must in A&D contracts, which is commonly done with 2D drawings. 3D LOTAR (Long-Term Archiving and Retrieval) standards are still under development. Users report that complex relationships between data and the physical products make implementing LOTAR difficult.
Leapfrogging, when one solution provider’s new features and functions start older PLM implementations on the downward spiral to legacy status.
“Breakage,” for want of a better term, when a core element of a PLM solution such as its browser version becomes obsolete. Microsoft Corp.’s Internet Explorer V6 is a recent example. Other obsolete browsers include Netscape Navigator, Mosaic 6, Prodigy Classic, MacWeb, IBM 0S/2 WebExplorer, and AOL Explorer. Whether any of these was used in a PLM solution is not known.

The remedies identified in the research are not one-to-one with the issues: they are on very different scales and do not line up neatly.

Typically, the issues are large in scale, widespread in the enterprise, and have several root causes. Many solutions still in use are built on old technology and implementation strategies that were not designed for sustainability. The remedies are either narrow and deep or broad and shallow, and either way don't completely solve any of the issues.

Good news comes on several fronts. New generations of technology are much more sustainable due to better underlying infrastructure and standards. Agile development processes are being applied to PLM implementations leading to less customization, which makes solutions easier to upgrade. Prominent among these is “wrap and reuse,” which is basically modular integration as an alternative to Rip & Replace.

More good news comes from DevOps (“development / operations”), the melding of application development with operations / production techniques such as Agile and Lean. Conceptually, DevOps lives at the intersection of software engineering, technology operations, and quality assurance. Most Cloud native solutions use DevOps concepts for solution updates that bring new features to customers incrementally with minimal disruption and a smaller risk of data loss. To say that the Cloud and DevOps are disruptive to old-style on-premise “monolithic” PLM is an understatement.

Perils in 'Customization' and 'Reverse Customization'

The obsolescence research reveals that some commonly proposed solutions haven't worked out well. Users constantly customize solutions, PLM and everything else, to enhance capability and productivity.

When the new software is deployed, data from the old version must be migrated in and solution providers strive to prevent incompatibilities. If customizations create or modify data in unanticipated ways, however, the data may be incompatible with the new version.

Solution providers continuously improve their product, add features and functions, and fix bugs, but users frustrate developers in when they:

Implement software and use it in ways developers didn't anticipate.
Skip updates and versions if the cost of disruption seems greater than the benefits.
Block specific functions in new releases due to issues in training, adoption, or data migration.
Keep functions in new versions looking like their predecessors, believing that this “reverse customization” is good for day-to-day productivity.

Don't see that solutions must meet the needs of all users in the enterprise, not just a vocal minority.

Longer term, of course, these users and their employers miss out on valuable new capabilities. Eventually these customizations have to be brought forward—another time-sink, a new batch of hidden costs, and more risk of data loss.

Preliminary Survey Results Uncover Hidden Problems

Preliminary survey results reveal that partnerships with software providers don't always work out well. Although often similar from one customer to the next, enhancement requests in A&D usually vary enough that solution providers implement them as customizations rather than as part of a scheduled product release.

Customization can add significant business value and yield a solid return on investment (ROI) in Product Lifecycle Management (PLM). Customization, however, usually involves modifying core capabilities, functions, and data formats. Because none of these is easy and any one of them makes upgrades difficult if not impossible, minimizing technology customization has long been a recommended best practice.

Nevertheless, for many reasons customizations are widespread. Some user explanations:

PLM solution providers' upgrades take too long and cost too much. Moreover, users expect to see their customizations—“enhancements”—in the developers' next releases. Users see these boosts in sustainability as carrying little cost and risk. Solution providers disagree.
The costs and risk of data loss in obsolescence are primarily technology based, i.e., a function of software coded by solution providers and controlled by them.
Solution providers have too much control over upgrades and how new solutions are to be implemented.
Most successful mitigations are primarily due to user skills and sound corporate governance.

Given that some PLM customization seems unavoidable, the CIMdata A&D Action Group is trying to get users to speak with one voice to resolve these issues.

Developing sustainable PLM solutions will free more resources for innovation instead of propping up outdated technology. For users and managements, PLM solutions should be kept as simple as possible, but no simpler.

The A&D PLM Action Group goal is to help solution providers develop technology that is sustainable and acts as insulation to offset “Third Rail” reluctance—the unwillingness to deal in a timely way with the costs and risks of technology obsolescence.

For some of those in charge of A&D's PLM solutions, technology obsolescence may seem like a Third Rail for some time to come.

Tom Gill is practice manager for CIMdata’s PLM Enterprise Value & Integration Knowledge Council and consults with industrial and solution providers. Mr. Gill worked 25 years in engineering and PLM within the automotive and industrial equipment segments. He holds a BS in Mechanical Engineering from the University of Maine.