KESTREL Swarm
Reduce self-sustaining congestion without assuming that one controller can safely optimise the whole system.
System context
A routing update increased average throughput but created severe intermittent jams at high utilisation.
Your role
You lead systems engineering for an automated warehouse with 3,200 mobile units.
Constraints
- The warehouse cannot stop for more than twenty minutes.
- Urgent orders must keep priority.
- A central controller would require new infrastructure.
Evidence pack
Local rule
A unit yields to a more heavily loaded neighbour.
Utilisation band
Jams appear only between 78% and 84% utilisation.
Priority effect
Urgent priority halves its own delay and doubles total congestion.
Simulation
Randomly slowing 4% of units nearly removes the jam.
Interactive simulation
Three connected decisions
Reduce self-sustaining congestion without assuming that one controller can safely optimise the whole system.
This case, its world, organisations, places and events are original Methodfield material and do not reproduce a protected fictional universe.