Case Study: Lockheed Martin Space Systems achieves rapid, accurate F‑35 fleet performance predictions and months‑faster simulations with MathWorks

Lockheed Martin Builds Discrete-Event Models to Predict F-35 Fleet Performance

Lockheed Martin Space Systems faced the challenge of predicting F‑35 fleet performance to minimize life‑cycle costs and maximize mission readiness across a complex global logistics system. Existing tools added complexity, so the team turned to MathWorks technologies—including Simulink, SimEvents, MATLAB, Parallel Computing Toolbox, MATLAB Distributed Computing Server, and Deep Learning Toolbox—to build a configurable, data‑driven simulation capability able to handle thousands of aircraft, parts, and scenarios.

Using MathWorks products, engineers built a discrete‑event Simulink/SimEvents model, ran thousands of Monte Carlo trials in parallel on a 256‑worker cluster with MATLAB Distributed Computing Server and Parallel Computing Toolbox, and used Deep Learning Toolbox to interpolate results. The MathWorks‑based solution cut simulation setup from months to hours, sped execution more than 20× versus desktop runs, reduced development effort, and substantially shortened overall simulation time, enabling faster, more accurate fleet‑performance predictions.

Lockheed Martin Space Systems

Justin Beales

Project Engineer