The two-level maintenance system presents some unanticipated challenges to maintainers. The author suggests that an evidence-based maintenance system would help them accurately diagnose faults.
As part of its transformation, the Army is converting from a four-level maintenance system to a two-level maintenance system. Unit and direct support maintenance assets are being combined to create the field maintenance level, and general support and depot maintenance are combining to form the sustainment maintenance level. For two-level maintenance to work, I believe the Army should develop a better approach to diagnosing faults. Two-level maintenance would be more effective if the Army had vehicles that were uncomplicated or were designed specifically for two-level maintenance.
Although vehicles have become more and more complex, the Army has fewer mechanics and a smaller logistics footprint. One would assume that enhanced onboard system-monitoring capabilities would increase the reliability of vehicles and reduce the time mechanics need to spend on vehicle repairs. However, based on what I have seen, I think the “failsafe” in Army maintenance has been the reliability of vehicles’ mechanical systems, not the electronic monitoring systems on those vehicles.
To take this a step further, think about the last time a malfunction light came on in your car. The fault likely was caused by a sensor that almost certainly was in error. Onboard diagnostic systems can tell you that a sensor is malfunctioning. Why can’t we develop a system that will predict when a sensor will fail? Why can’t we determine (with a relatively small deviation) how often that sensor fails? Perhaps we need to develop a built-in sensor test, design around that sensor, or eliminate it.
All too often we seem to believe that adding sensors makes the hardware we are given the best available. I find that prognostics on ground vehicles are generally complex and cause more problems than they solve. This observation conflicts directly with the direction the Army is taking with ordnance military operational specialty skill sets. Each maintainer is now tasked to know multiple vehicle systems and repair systems. However, maintainers will probably never learn each system well enough to become proficient in troubleshooting them all.
In healthcare, Internet-based systems are available to help doctors identify possible causes for patient symptoms. One such statistical diagnostic assistant, called “Isabel,” was developed by a father who sought to change the diagnostic system that affected the way his daughter (Isabel) was treated. This system is basically an intuitive system that takes advantage of all previous diagnoses and provides the statistically most likely disease (fault) and treatment (repair).
|The high-mobility artillery rocket system consists of a missile
system and the truck that transports it. Both must be well-maintained and operational for the system to
The medical profession has done well to embrace evidence-based medicine; the Army maintenance community also might do well to embrace evidence-based maintenance. Not only would such a system speed troubleshooting and proper fault diagnosis, it would also perform the same function as prognostics by supporting the building of better authorized stockage lists. It would also provide a better grasp of whole-life and life-cycle costs. A system like this for Army maintenance could limit misdiagnosis of vehicle faults. An evidence-based maintenance system would reduce the multicapable maintainer’s reliance on his intuition to make sound repair decisions and allow him to diagnose problems and correct them. His data then could be included in later regressions to ensure that faults were categorized correctly.
The program should be able to determine the result statistically. For instance, if the maintainers normally say that a certain fault is found during testing, we could probably determine what the true fault usually is (and more importantly, what the fix is) or if the “how found” data mean nothing to the outcome. I would say, at this point, that we do not really know if these data mean anything, because we have “intuitively” said they mean nothing.
The most intriguing benefit of intelligent fault diagnosis is that it eliminates the need for the intuition of the maintainer. A maintainer’s intuition results from his general maintenance experience, training, and experience with the given piece of equipment. We need to find a way to capture the shop data and fault data from the Standard Army Maintenance System Enhanced or from a Department of the Army Form 2407, Maintenance Request, database that captures the “how found” data and ties them directly to the “how repaired” data.
With the advances in controller area network bus technology over the past 10 years, several large commercial truckers have developed evidence-based service regimens based on what they have learned from their electronic monitoring systems. The Army needs similar systems that will support “fight with what you brought” because new prognostic-laden equipment and smart vehicle systems will not be widely used for years to come, but evidence-based maintenance is available today. It also would be free because the data are already there—just not being used. Worse yet, we are losing the information by not properly archiving it.
We have all the tools needed to use evidence-based maintenance without adding anything to the vehicle systems in current inventory. The type of bus used on virtually all heavy vehicle systems has the data we need and stores them quite accurately, but we rarely interrogate it.
One might do well to study exactly what commercial, even consumer, products provide. OnStar offers a level of condition-based maintenance to the consumer. My truck lets me know, based on my driving habits over the past few months, that it will need an oil change soon. The onboard computer sends a message that generates an email telling me I am down to a percentage of oil life remaining. Lo and behold, a few days later my service engine light illuminates. Granted, all of this is done with the assistance of a few sensors, but I think more weight is given to the “profile” of miles driven (such as revolutions per minute duration) than to the oil condition itself.
Time and time again, I read on LOGNet about the need for simpler vehicles that align with the workload and skill set of the Army maintainer. Perhaps the greatest single issue that comes up is the complexity of maintaining a central tire inflation system (CTIS). Soldiers in the field seem to be content with vehicles that do not have CTIS or that have disabled CTIS systems. CTIS is a complex system, and many fleets still do not have it. I am not privy to the results of the surveys that take place, but I think units in the field have overwhelmingly said they do not want CTIS systems since any prognostics undoubtedly will increase the cost of the vehicles and make them inherently more difficult to maintain. Maintainers do not want to have more complicated vehicles unless they have the proper tools and adequate knowledge and understanding of the vehicle system. A proper tool, in this case, would be a diagnostic system that has the information needed to diagnose a fault properly with a high degree of accuracy.
Anything less than having a high degree of accuracy in troubleshooting diagnoses equates to simply changing parts. If we know more about the predicted actual fault, we can eliminate some of the practice of “changing parts until the fault goes away.” More importantly, we might be able to use the data to redesign our resupply operations. To shrink the logistics footprint, we need to do several things. If we are not going to make vehicles simple, we need to make diagnosis simpler. We can do it without adding anything to the ground fleet.
Staff Sergeant Michael Winkler is a senior editor and course writer for the Leadership Development Directorate, 84th Army Reserve Readiness Training Command, at Fort McCoy, Wisconsin. He holds a bachelor of business administration degree from Marian College and a master of business administration degree from Indiana Wesleyan University.