VIEWPOINT DEFENSE DEPARTMENT
Days of Cannibalizing Military Aircraft for Spare Parts Coming to an End
Photo: Air Force
Since their advent in the 19th century, interchangeable parts have allowed manufacturers to efficiently produce products in quantity and for lower costs. And after a product leaves the manufacturer’s assembly line, such parts make repairs significantly easier than handmade or custom-fitted parts ever allowed.
Interchangeable parts are undoubtedly a fundamental building block of success in industrial manufacturing.
But today, as mechanized products evolve into an element of a more integrated complex system, changing out a failing part for an identical interchangeable part may not be as straightforward as it has been in the past. In defense, especially, as state-of-the-art weapon systems advance to include architecture that exploits automated self-healing recovery and artificial intelligence, swapping out parts may no longer be possible without jeopardizing the delicate relationships among linked components and subsystems within the greater system.
For many years, perhaps decades now, maintenance crews realized that if they had two or more similar complex systems, say, frontline military aircraft, and more than one was down for different maintenance reasons, they could remove parts or components from one system to repair the others. In many situations, cannibalizing one system for its interchangeable parts enabled maintainers to return failed systems to working order more quickly than did troubleshooting and repairing individual problems or diving into inventory, finding the appropriate new part, and installing it.
The reasons for cannibalism are numerous. If a part had to be shipped from a depot, the original manufacturer or a subcontractor, the wait time could be excessive. Furthermore, if the user community had not paid for availability, the part may not be sitting on the manufacturer’s shelf available for shipping. When budgets are reduced, such as with sequestration, parts availability is one area that is quickly underfunded. So, “living like a cannibal” evolved into a procedure of necessity for maintenance crews.
The consensus opinion is that underfunded spares may be the principal factor in this now well-established procedure of cannibalizing one aircraft to preserve readiness in others. The Government Accountability Office and other investigators have highlighted the possible causes and ill effects this practice creates.
Recent investigative reports in the news focus on readiness rates that are in some cases below 50 percent — a startling reality. To improve woeful readiness numbers, maintainers select a donor aircraft from which to secure parts — parts that are otherwise not available. They source the donor aircraft repeatedly to preserve airworthy status of other aircraft. A cannibalized aircraft may sit on a flight line for months, and when an effort is later initiated to restore the donor to mission-ready status, it is a monumental task. Regardless of the reason or the cause for cannibalizing complex systems, the era of interchangeable parts and cannibalism is coming to an end.
To some degree, today frontline fixed-wing and rotary aircraft all have built-in test and self-diagnostics that develop a characterization of the unique overall system. The diagnostics sample the original components and subsystems as they bond and couple through use and wear over time. The internal components continuously poll and mutually sense each other’s performance under various operational conditions. They also develop distinctive characteristics based on their parts dependencies, subsystem relationships and operational environment. In complex systems, the collective behavior of components goes well beyond the anticipated response from the sum of the parts.
For these state-of-the-art systems to reach the user community, they must pass manufacturers’ rigorous static testing, such as vibration, temperature, pressure, electromechanical interference, and burn-in testing, and operational testing in anticipated field environments. With each acceptance test, complex systems develop even greater cognizance of their own embedded relationships among parts, components and subsystems.
Whereas many believe artificial intelligence pertains principally to external relationships — for example, interaction between a human operator and an unmanned aircraft or a driverless vehicle sensing its surroundings and maneuvering accordingly — internal functionality of today’s complex systems requires an element of embedded intelligence to poll components and rapidly optimize internal system performance.
AI raises the level and sensitivity of interfacing because it can process terabytes of data. State-of-the-art systems efficiently perform predictive analysis, detect faults, and reroute and reconfigure through handshakes that couple components and sense thermal and electrical parameters, radiation and pressure, among other variables, to ensure uninterrupted operation.
Complex systems with AI capabilities learn how to optimize internal performance using functional analytics and data exchange among components. Cycle time for adjusting component and subsystem parameters to enhance performance is measured in microseconds — beyond a threshold where human control is possible. The algorithms, logic, processing, and convolutional computing in AI-embedded systems enable these systems to operate virtually independently of human interface.
Today’s military aircraft — the very definition of complex systems — are extremely sensitive to changes. An AI-enhanced system may perceive any variance in the data stream from coupled components as an anomaly or, worse, a fault that endangers the larger system. If a component or subsystem is removed and replaced with another — even one manufactured by the same company on the same line — the part may vary ever so slightly within specifications that it jeopardizes aircraft performance. A simple cannibalized parts swap bypasses the original coupling of components and characterization testing and may very well jeopardize system integrity. Interchanging parts indiscreetly, without full consideration of the impact on the larger system, is not advised.
As systems become even more complex, however, maintainers increasingly will have difficulty diagnosing problems, and test equipment on the flight line and in the back shops will lack the sophistication necessary to identify failing or failed components, which will both result in a greater reliance on cannibalization. But a narrow focus on a particular part or component replacement may in turn impair performance of the greater system. Pressure from field commanders to maintain the highest possible readiness status through cannibalization must be met with resistance — repair versus replace should become the norm.
To move forward, training for maintenance personnel must advance as rapidly as the systems they are being asked to maintain. Their tool chest of test equipment must match that of the original manufacturer, and the data flowing through the internal raceway must be captured and analyzed holistically before certifying airworthiness and mission readiness.
Full-system run time must be sufficient to ensure coupling, characterization and anticipatory output from replaced components. The system must provide indications so maintenance crews know when full system integrity has been restored. Training course curricula must be developed in partnership with manufacturers to ensure maintainers appreciate the synergistic impact of each component and subsystem.
Over the past several years the services have experienced a spike in aircraft accidents; according to Military Times, there has been a 40 percent increase. Much of this can be contributed to the lack of sufficient flying hours for training. Funding shortfalls have been the cause of reduced flying hours, as well as an increase in cannibalization. The F-22, F-18E and F-35 are by their very nature extremely complex sophisticated systems; therefore, singular parts cannibalization without proper full system coupling may be a factor in these accidents — a question that must be explored.
As tomorrow’s complex systems embrace intelligent technology and develop greater ability to sense and adjust to internal and external variations for mission success, we must move away from a component focus and instead acknowledge the holistic nature of larger AI-embedded systems.
As Frank Michael and Shane Shaneman stated in the August “NDIA Perspectives” column in National Defense, the imperative to preserving battlefield advantage is cooperative development among government, industry and academia. These three entities must address in earnest discussion the implication associated with cannibalization of today and tomorrow’s sophisticated complex AI-embedded systems. The current development path may very well bring an end to “living like a cannibal.”
Retired Air Force. Col. John C. Johnson is a former vice president and general manager at Northrop Grumman Corp. Contact him at email@example.com, or sandvistallc.com.