You'll never walk alone - simulating pedestrians in a world of cars
summary
Demonstrates a working JuPedSim+SUMO coupling that runs pedestrians and cars on the same microscopic timeline, importing OSM, enriching it (guessed traffic lights and lane connections), and placing flows such as an 1,800 pedestrians flow on sidewalks. Uses the osm web wizard to generate scenarios, netedit to add demand, and the TraCI Python control to pause a person (ID 0) mid-simulation. Claims integrated models catch crowd effects for safety, runs as C++ libraries with nightly ~15,000 tests and open source tooling.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 6:46Why walking matters
- 15:19Pedestrian movement models
- 21:32SUMO scenario setup
- 27:23First pedestrian model
- 31:44Editing pedestrian demand
- 34:01Programmatic pedestrian control
- 36:10Testing and improvements
- 37:39Applications and future
related talks
29mDecentralized Signal Control for Urban Road Networks
22m"How We Built a Rust-y Real-Time Public Transport Map"
51mBuilding concurrent systems using Akka.NET
54mMicro-Mind Services: Building Complex Behavioral Systems
14mConnecting Reinforcement Learning to Simulation Software
35mIt’s An Evolution: Recreating Life With Elixir Processes
34mSimulating Our Apache Mesos Framework for Fun and Profit and...
52mHow to Run a Metropolis Without SMS or Registration: The Story of an ASUDD Pilot