Defense logistics
Improving logistics and maintenance decisions for a defense organization
Quantitative support for logistics and maintenance, repair, and overhaul decisions where availability matters more than cost alone.
Client
U.S. Department of Defense organization
Basis
Partner experience, prior role
The insight
“When a shortage is measured in lost availability rather than lost sales, retail inventory rules point in the wrong direction.”
Exhibit
Rule-based stocking sits inside the frontier
Asset availability achievable for each level of spare-parts investment
Schematic, not client data. The frontier is the best availability any stocking policy can achieve at each budget. Moving to it means more readiness for the same money, or the same readiness for less.
Readiness depends on the right part in the right place
The organization supported maintenance, repair, and overhaul operations through a complex logistics network, where the availability of the right part at the right place directly affected the readiness of the assets it served.
Retail rules do not fit failure-driven demand
Demand for repair parts is intermittent and driven by failures that are inherently uncertain. Lead times are long and variable, and the cost of a shortage is measured in lost availability rather than lost sales. Rules of thumb built for commercial retail inventory do not transfer well to this setting.
Optimizing for availability, not turns
We applied probabilistic modeling and optimization to logistics and maintenance decisions, framing the objective around availability and readiness rather than inventory turns. Simulation was used to test policies against realistic variability before recommending changes.
A defensible basis for logistics decisions
The work gave decision-makers a quantitative, defensible basis for logistics and maintenance decisions. Specific outcomes are not published, consistent with the nature of the engagement.
What made the difference
01
Model failure-driven demand
Parts demand was treated as intermittent and uncertain, driven by failures rather than by a steady sales rate.
02
Optimize for availability
The objective was framed around readiness of the assets served, not inventory turns or cost alone.
03
Test before recommending
Simulation tested candidate policies against realistic variability before any change was recommended.