EfaSim - A Prototype for a SUMO-driven Emergency Vehicle Driving Simulator
summary
Prototype named EfaSim integrates SUMO for traffic and Unreal Engine for rendering, using a custom DLL via Libsumo instead of TraCI websockets to reduce latency. Implementation spawns/despawns vehicles from a pooled cache, represents the ego vehicle as three SUMO agents for realistic interactions, and benchmarks frame rate and latency. Results show similar FPS for both methods but lower latency and better information throughput with the DLL, able to simulate 220 cars at 60 fps.
AI-written summary · about
chapters (11)
- 0:00Introduction and goals
- 0:53Unreal and SUMO integration
- 2:05Architecture and data flow
- 4:12Simulation update process
- 6:34Ego vehicle and loop logic
- 9:24Simulation features
- 11:12Pipeline and environment
- 12:15Performance results
- 14:10Latency analysis
- 16:08Conclusion and future work
- 17:52Final remarks
related talks
22mA Fully Immersive Driving Simulator for Autonomous Vehicles
32mGamification of Transport and Infrastructure Engineering
14mCross-Car, Multiplayer Games for Semi-Autonomous Driving
05mMixed Reality Visualization in Automotive SDR Networks
11mMaximizing Efficiency and Minimizing Costs with Unreal Engine
10mNative Parity for Development & Testing of SOAFEE Software-Defined Vehicle
33meSim developed by FOSSEE
28mOpenAdx "Autonomous Driving Simulation" Testbed – hardware in the loop