e-CM: Electronic Collaborative Maintenance for Depot Repairs and Manufacturing
DoD Participants: U.S. Army – Tobyhanna Army Depot
Obsolete or hard-to-find parts are a problem that plagues all Maintenance Depots in all branches of the military services. Electronic parts are particularly troubling. The well-known Moore’s Law states that integrated circuit capabilities (component density and clock speed) double about every 18 months. As a result, complex semiconductor integrated circuit designs such as microprocessors, memory and modem chips have limited salable lifetimes – usually well under 10 years. Application-specific integrated circuits (ASIC) or the custom-designed integrated circuits frequently found in defense systems are not exempt. Semiconductor manufacturers scrap or sell process equipment when the manufacturing process advances render an old fabrication plant uneconomical to run, and “old” usually means less than a decade. With defense electronic equipment lifetimes of 20 years or more, almost all such systems will, before the end of their useful life, encounter a situation where a component needed for repair is no longer available from routine stock.
Before a semiconductor manufacturer ceases production of a given component, known users are offered the opportunity to make a “lifetime buy,” or to purchase and store a quantity considered sufficient to meet all future requirements. That is always the case for ASICs. Clever component suppliers recognizing that a lifetime buy of components present significant economic opportunities, purchase them to make them available at a demand-driven price. Furthermore, enterprises have emerged that specialize in reverse engineering ASICs to make replacement parts. Either source could offer a solution for hard-to-find parts. The problem for Depots that perform electronic maintenance is how to locate them.
The current solution is for estimators to spend hours, even days, telephoning potential suppliers. This project sought to develop a web site to which Tobyhanna (TYAD) estimators could post lists of parts sought, listed by project, and broadcast notifications of new posts to a list of potential suppliers. Suppliers could then scan the parts list and enter budgetary estimates of part prices for any parts they could supply. The estimator could then provide a much more accurate and timely estimate of the cost to repair a given system. TYAD procurement could use the budgetary price responses to create Request for Quote (RFQ) lists. The overall result would be greater productivity for all – estimators, suppliers, and TYAD procurement.
This project tried a new approach with a minimal project team where the matching funds from industry were to be documented from the web site itself by collecting metrics on industry usage during the pilot trial. BRL, Incorporated created the web site using requirements provided by TYAD. BRL hosted the site during the pilot trial, and TYAD was to host it for production operation.
Initial development efforts proceeded on plan except for minor delays encountered in gathering sufficiently detailed user requirements to enable software development. Design reviews during development were problematic because it proved very difficult to capture enough dedicated time with the very busy estimators. Shallow design reviews created problems farther down the development path.
An initial list of suppliers was provided by TYAD. The National Center for Manufacturing Sciences (NCMS) validated contact information and populated the initial supplier database from that list. The initial list of estimators was populated from data gathered early in the process.
The project went on-line for trial use and those shallow design reviews proved harmful. Issues were identified that should have been caught before going on-line. Actual system software bugs were few and easy to fix, but ease of use issues proved serious. For example, one issue that in retrospect should have been obvious but was not at the time was the fact that suppliers could only enter estimates for parts within the bounds of a single project. It would have made more sense to combine all like parts across project boundaries with a single estimate for all. Both estimators and suppliers found these ease of use issues serious enough to limit their willingness to use the system.
At this point the project funds had been exhausted. The small company that created the system was in no position to offer additional work as in-kind and NCMS had no discretionary funds to invest in further system development. Small functional fixes were implemented but none of the major corrections could be made pending identification of additional funds.
BRL hosted the system at first but as usage dwindled toward zero, NCMS moved the servers to its facilities and kept the system on-line for several months.
With corrections remaining to be implemented and no funds located, the sole remaining option to continue the work was for TYAD to begin paying a subscription and system maintenance fee. Further complicating the work was Army server reduction efforts. The original plan had been for TYAD to host the system, but that was no longer an option. After several months of fruitless negotiation and searching for an alternative funding source, NCMS reluctantly took the site down.
Significant potential benefits in terms of increased productivity for both TYAD estimators and their suppliers were identified but never realized. That means that an opportunity was lost. Loss of opportunity is always painful but lessons learned here can be used to avoid future issues.
· The first lesson concerns industrial matching funds. This project planned an innovative means of capturing matching funds. During the trial period for the system, access logs would capture who used the system and for what purpose. Metrics would also be captured on log-in time. Using assumptions on hourly cost of different resources, the number of log-ins and the average time on-line would yield industry’s system usage costs and that would be logged as industry match. Since system usage was minimal during the trial period, matching funds were also minimal. Recommendation #1 – the 2/3 project cost industry matching funds should always take the form of contractually committed in-kind. In-kind above the 2/3 level can be of softer commitment.
· Software projects can be difficult because software product development cycles are in many cases shorter than the time required for forming and launching a collaborative project. However, there are many different domains of software products, some are appropriate for Commercial Technologies Maintenance Activities (CTMA) projects while others are not. General information infrastructure is the domain of network communications hardware/software and enterprise application software. Vendors are mostly large companies such as Cisco Systems, SAP, and so forth. Competition is fierce and development cycles short so that collaborations will almost always be behind the technology curve. There are few opportunities for CTMA projects here. Recommendation #2 – projects targeting infrastructure and/or enterprise applications should generally be avoided. Occasional, highly focused exceptions may arise but are likely rare.
· Software tools focused on manufacturing productivity, cost reduction, or part quality tend to be on longer development cycles. CTMA projects in this domain can be and have been very effective for both original equipment manufacturers (OEMs) and Maintenance Depots. But even in this fruitful domain caution is indicated. Good ideas frequently come from small start-up companies but inability to provide matching funds at some level or lack of financial backing for a reasonable product launch should raise caution flags. Recommendation #3 – software tools focused on manufacturing productivity, cost reduction, or part quality should be considered appropriate targets for CTMA projects but caution is proper.
· Caution #1 – smaller technology providers should be willing to offer at least overhead costs as matching funds, or the project should include one or more larger companies (perhaps large technology users) that will commit to provide sufficient matching funds for the project to meet the 1:2 model project target.
· Caution #2 – software tools that offer a niche solution in a wider market segment should have strong Depot support and strong commercialization interest by a major technology provider.
· Caution #3 – web-enabled tools are great, but commitment for long-term software support by a commercial host is essential.