Warehouse Robot Traffic Deconfliction

Fleet Traffic

Robots per fleet 10

Simulation speed 1.0×

Live comparison

Uncoordinated

Deliveries

–

Per minute

–

Jammed now

–

Total jam-ticks

–

Traffic-managed

Deliveries

–

Per minute

–

Waits inserted

–

Reroutes

–

Throughput over time

How it works

Both warehouses are identical grids of shelving aisles. Every robot computes its own shortest grid path (A*) to a randomly assigned pick location. The left fleet simply follows that path — when two robots' shortest paths need the same cell at the same moment, neither yields, so they lock up and the jam spreads backward through the aisle behind them.

The right fleet adds a priority-based traffic layer: whenever a robot needs a new route it plans against a shared reservation table of every other robot's near-future cells and time-steps (a space-time A*), inserting a short wait or a slightly longer detour so its path never collides with an already-reserved cell. No central robot is special — priority is simply "whoever reserves first, blocks the rest" — yet the fleet keeps flowing.

Robots per fleet — raise it to see uncoordinated throughput collapse toward zero at busy intersections while managed throughput degrades much more gently.
Deliveries/min — a rolling rate computed from completed pick-ups in the last simulated minute, for direct side-by-side comparison.
Jam-ticks — how many robot-ticks were spent stuck nose-to-tail in the uncoordinated fleet; the managed fleet instead accumulates small planned waits and occasional reroutes, which is why it stays productive.

UncoordinatedTraffic-managed

Shelf / rack Pick location Jammed robot Reserved-route robot

⚠ Couldn't load the 3D engineThree.js failed to load from the CDN. Check your connection and reload.