Physics Simulator for Bin Optimization

This precision machine shop produces thousands of irregularly-shaped metal parts daily. Workers packed bins by instinct—often underfilling to avoid jams or overfilling and causing damage. The result: 20% of bin capacity went unused, shipping costs ballooned, and inventory forecasts were unreliable. Management had no way to predict how many bins a production run would require.

We developed a custom physics engine that simulates gravity, collision, and settling behavior for parts of any shape. The system ingests CAD models or scanned geometries, runs thousands of virtual packing iterations in seconds, and outputs optimal bin configurations. Workers now see exactly how to load each bin before touching a single part.

Bin utilization jumped from ~65% to over 85%. The shop reduced shipping frequency, cut packaging material costs, and gained predictable inventory planning. What once required hours of trial-and-error now takes seconds of simulation.