When establishing an integrated, effective, affordable sustainment organization, you start with a good sustainment risk process. Then you move on to establishing an effective and efficient system observation process using these 5 steps:
- Use your “free” data
- Look to your repair depots
- Set up an age surveillance program
- Establish processes for special testing
- Analyze your data to create information
Here are some words on step 2:
Repair shops, depots, and other similar facilities are often created and run by organizations purposefully focused on productivity and throughput in order to ensure economical repairs. A great example of this tendency is the Air Force Sustainment Center “Art of the Possible” e-book[i] which is replete with ways to ensure economical production via speed. (example: “Chapter 1 Introduction: The Value of Speed”) This is exactly what the Sustainment Center should be doing, saving tax dollars via efficient production.
However, if you are not careful, this approach can be sub-optimal to your goals of observing your system to identify and mitigate risks with sufficient lead time. For example, they may be focused on rapid remove and replace processes where the original failure can be lost in the shuffle. See figure 3. Therefore, your sustainment organization will need to contract with your repair depots to ensure you get the data you need, when you need it. The best scheme for this is the Closed Loop Failure Analysis (CLFA) program.
CLFA is MIL-HDBK-2155 FRACAS (Failure Reporting and Corrective Action System) employed at a repair depot. FRACAS is designed to discover problems during production by finding unforeseen failure modes. It can be used very effectively to find new emerging failure modes during the sustainment phase.
CLFA seeks to “close the loop” between the failure noted and the repair made. In its most basic form, it does this by asking the question in the top left of the graphic above: “Did this bad part create this fielded failure?”
An affordable repair depot works to increase throughput while keeping costs down. Diagnosis is performed to the extent needed to ensure a repair. Large depots can contain many shops that perform their work somewhat independently of each other. A subsystem might be delivered to the depot and have a component removed and replaced and the subsystem is tested and sent out again. Meanwhile, the component may have a part removed and replaced and then the component is tested and placed on a shelf as a good spare.
This remove and replace strategy is a good one for speed, throughput, and affordable repair. It will certainly be the model used for in-orbit depot repairs. It has the potential to allow failed parts or components to remain in service and for emerging failure modes (those not thought of during design) to remain undetected.
Good sustainment organizations will enter into agreements with their depots to ensure repairs are traced back to failures and sufficient diagnostics are performed to ensure emerging failure modes are found. This sounds expensive, but even if the program starts with 100% screening, it can quickly eliminate from scrutiny those failure modes and repairs that are well known or become well known after CLFA has functioned for a few years.
Having proven itself, a mature CLFA program will not limit its investigations to only system operational failures. Subsystems and components can and do fail in the depot and elsewhere, revealing important system assessment data. In addition, a mature CLFA program will also take the findings of its failure review board and use them to improve its own depot processes and equipment.
CLFA demands good processes and a good information management system to handle the large amounts of data generated and analyzed.
There is also a clear benefit to the repair depot. More intense tracking will prevent poor components from circulating between customers and the depot. This happens nowadays when new emerging intermittent faults combine with outdated diagnostic and final testing. This leads to “bad actor” systems being released to the field, often, but not always, the same serialized box. The possibility of this “bad actor” syndrome should be considered when failures of a particular subsystem begin to increase.
[i] “Art of the Possible”, 367TRSS.Hill.AF.Mil, 20 July 2015