To be cleared for flight, F-35 policy states that an aircraft must be electronically “complete” in ALIS [autonomic logistics information system], meaning that all of the electronic records from each installed F35 part must be entered into ALIS. However, users at all 5 of the locations we visited told us that electronic records are frequently incorrect, corrupt, or missing, resulting in ALIS signaling that the aircraft should be grounded, often in cases where maintainers know that the parts have been correctly installed and are safe for flight. Users at 1 location said that within a 6-month period in 2019, they experienced anywhere between 0 and 400 issues per week related to inaccurate or missing electronic records. These same users said that it is common for their squadron leadership to elect to allow an aircraft to fly with over 20 inaccurate or missing electronic records that ALIS signals to ground.
That was from an informative GAO report, “DOD Needs a Strategy for ReDesigning the F-35’s Central Logistics System.” Anyone paying attention knows that the F-35’s ALIS has had a long history of trouble. The number of open deficiencies continues to grow steadily and is around 4,700 in Sept. 2019. That same month, officials decided to stop releasing major upgrades to ALIS.
ALIS was first developed in 2001. That’s not before the existence of cloud technology, but it was well before Amazon AWS existed and the start of wave cloud/mobile-native successes.
There are three related efforts for developing the next generation of ALIS software, shown below. Note the relatively small sums being expended — less than $75 million through FY 2019 — but I presume as legacy ALIS ramps down these numbers will rise fast.
This next part is super interesting — the culture clash between (1) those advocating for a government in-house enterprise tools architecture and (2) those looking for a more traditional outsourcing to commercial solutions:
DOD officials we spoke with expressed differing views on the extent to which DOD should adopt an Agile software delivery model for ALIS. For example, in a 2018 memorandum establishing the Mad Hatter pilot, a senior Air Force acquisition official stated that the F-35 program should embrace the tenets of this type of model in order to innovate and rapidly deliver useful capability through ALIS. Similarly, Air Force, Office of the Secretary of Defense, and some F-35 program office officials stated that modernizing ALIS will require DOD to adopt industry best practices by making decisions quickly, delivering usable products early and often, and revising plans to reflect experience from completed software iterations.
In contrast, Marine Corps and some F-35 program office officials indicated that DOD should carefully consider different commercially-available software tools, as well as DOD-specific constraints, before delivering new ALIS capabilities. For example, F-35 program office officials associated with the ALIS Next initiative stated that they conducted an assessment of the commercial software tools that could be used for new ALIS software development. These officials said that some of the tools that were initially being used by the Mad Hatter initiative to develop applications make software development easier in the short-term but more difficult to switch toolsets and/or contractors in the long-term. Marine Corps and some F-35 program officials also noted that current DOD processes and procedures—such as the software certification and cost-estimating processes—may not be able to support quick software releases. While an Agile software delivery model has been identified as having the potential to improve the way in which the federal government develops and implements IT, we previously reported that this type of model requires significant procedural and organizational changes in order to be implemented successfully.
… DOD has not made a decision about the extent to which the ALIS redesign will be hosted in the cloud as opposed to onsite servers at the squadron level…. While the Mad Hatter initiative has embraced hosting ALIS in the cloud, including at the squadron level, ALIS Next is conducting an assessment of the extent to which a cloud-based system is the best option for ALIS.
So it turns out the F-35 program office may not move onto a cloud-based solution! But the F-35 stops major development on ALIS soon as it moves to a Government-led Operational Data Integrated Network (ODIN) supported by Kessel Run, Lockheed, and others like the 309th out of Hill AFB. Fall of 2021 is when the Air Force wants to switch a squadron over from ALIS to ODIN.
I wonder if the question of cloud-based solution or not has been put to bed yet. One of the big questions here is the F-35’s adoption of the Air Force’s enterprise software tools, and the entire agile/devsecops philosophy. Perhaps you can be for agile/devsecops and also think the Air Force’s toolset is too prescriptive. But unless there’s some magic bullet I haven’t heard about, I why not give them a shot? Architect it the way the Air Force Chief Security Office Nicholas Chaillan recommends, and then run with rapid releases.
Perhaps this is a good opportunity to run an experiment. Let Kessel Run and the Air Force CSO quarterback on one development stream and then also pursue a commercial product the usual way. Both would likely involve Lockheed Martin to fair extent. With ALIS ramping down it could help fund the two projects. In any case, to those who argue it is too expensive, the last 19 years would indicate that it was too expensive to attempt a single-best solution!
Leave a Reply