A Swarm of Processes — Simulating Ant Foraging Behavior...
summary
Simulation claims a full foraging colony using a separate OTP process per ant, with per-ant GenServer actors, a DynamicSupervisor + Registry via tuple naming scheme and 400 tiles as tile GenServers on a 20×20 grid. Each turn runs synchronously (about 6 ms per turn, UI updates every 50 ms), hundreds of ants coordinate via pheromone values on tiles, faults restart errant ants, and surprisingly overall runs faster when concurrency is reduced rather than maximized.
AI-written summary · about
chapters (9)
- 0:00Start
- 5:32Ant System
- 6:44Inclining Optimization
- 10:41Data Types
- 10:56Define Our Humble Ant
- 14:29Supervision Tree
- 14:55Dynamic Supervisors
- 16:32Simulation Supervisors
- 21:30Fault Tolerance
related talks
57mWhat is Swarm Intelligence? - Swarm AI Explained
52mAn Introduction to Swarm Intelligence Algorithms
42m¡Viva La Evolución! Replicating Life with OTP
30mStupid Ideas For Many Computers
16mRats Of The Maze
35mCellular Automata How to become an artist with a few lines
33mDeep Space - A Simulation Drama
01h 00mThe Forgotten AI of Video Games