Where’s a spare part when you need it? For the Department of Defense, that’s the multibillion-dollar question looming over the F-35 Joint Strike Fighter program.

According to the Government Accountability Office (GAO), 30 percent of F-35 aircraft were grounded awaiting replacement parts between May and November 2018 — three times the accepted program target rate. “Specifically,” GAO’s April 2019 report states, “the F-35 supply chain does not have enough spare parts available to keep aircraft flying enough of the time necessary to meet warfighter requirements.”

This situation can be traced to the predictive modeling that underlies how DoD and original equipment manufacturers (OEMs) calculate target availability for spare parts. These models are fed forecasts that, while trustworthy in some contexts, are ill-equipped for the volatility that effects DoD operations. Contingency operations, continuing resolutions, and disruptions within the global supply chain (such as Turkey’s tenuous supplier status) unduly impact the demand for parts — sending the most sophisticated forecasts askew.

Unstable demand puts OEMs in a position of producing parts only when other, more certain production priorities can accommodate it. This leaves DoD with lead times of months or even years for certain parts. Long lead times make forecasts even more unreliable.

The F-35 program’s shortcomings in this regard have been observed before. In our support to the Defense Logistics Agency (DLA) and service branches over the past decade, we found these models are subject to large errors, both in evaluating the tradeoff between wait time to issue parts against cost and in the recommended mix of spare parts to attain specific aircraft availability targets.

We found a better way forward. In an uncertain world, it’s best to hedge your bets and remove forecasting from the equation. Use a robust strategy that will likely lead to a good outcome, whatever may happen.

Advanced mathematical algorithms now make such a strategy possible. Forgoing forecasts and variances as inputs, these hedging algorithms use transactional data to produce control levels that generate a better mix of spare parts regardless of the demand scenario. This means more of the parts requested by maintenance are available when needed and fewer parts sit on the shelf.

DLA implemented hedging models to great success six years ago: parts availability on frequently demanded items jumped from 85 to 90 percent. The increase was even greater for less frequently demanded items, from 72 to 82 percent. If the F-35 program were to see a similar boost in parts availability, far fewer than 30 percent of aircraft would be grounded for parts — perhaps even below the 10 percent target rate.

Our analysis suggests supply chain hedging could increase aircraft availability across the F-22, F-16, and F-18 programs as well — putting DoD’s stated goal of an 80-percent mission-capable fleet this fiscal year within reach. Aircraft availability relies on support to bases as well as the depots where major repairs and procurement take place. Hedging algorithms enabled DLA to balance its investment appropriately between these essential locations.

The cost-efficiency implications are profound. DLA has achieved a recurring cost avoidance savings of $400 million annually, on top of a one-time, $600 million inventory cost reduction. DLA’s experience foreshadows a significant opportunity for the F-35 program, which is projected by GAO to incur $1 trillion in lifecycle sustainment costs.

DLA’s success has spurred DoD and the U.S. Air Force to pursue hedging models as a means to reduce inventory costs, but it has yet to become the norm. Success within the F-35 program could change that. Certainly the potential for benefits to be realized across DoD exists, as well as support to OEMs and performance-based logistics outcomes.

Hedging models help decision makers see the tradeoff between key interests (e.g., inventory value, average wait time for parts, total annual buy and repair spending) and set a path that meets multiple potential demand scenarios — enabling DoD to be better prepared for the unpredictable.

Supply chain hedging may not be a quick fix to the F-35 program, but with the prospect of greater fleet availability and lower supply chain costs, it could be a lasting one.

Tovey Bachman, PhD, has more than 30 years of experience in operations research and DoD logistics. He is lead developer of the Peak/Next Gen (PNGTM) algorithm development team at LMI, a not-for-profit government consultancy based in Tysons, Va.

Colonel (ret) Pat Kumashiro is a director at LMI. He served as a maintenance and logistics officer for 27 years, including 4 years as the HQ USAF Maintenance Division Chief. He is a former director of the Eisenhower School Supply Chain Management Concentration Program at National Defense University.