AutoSim Pro
Autonomous vehicle simulation platform with vehicle physics, AI behaviours and analytics
- Status
- Research
- Version
- No tagged release
- Platform
- Desktop app
- License
- MIT
Gives researchers, engineers and students a controlled environment to test and study autonomous driving algorithms and vehicle behaviour.
How it is put together.
AutoSim Pro is a desktop autonomous vehicle simulator written in Python with a PyQt6 interface and OpenGL 3D viewport. It simulates multiple vehicles of seven types with a physics engine (including a simplified Pacejka tyre model), configurable SAE autonomy levels, behaviour trees and neural-network behaviours, weather and lighting, traffic lights and pedestrians. It includes an analytics dashboard with CSV/JSON export, recording and replay, and scenario, behaviour and map editors.
- 1
Simulation core
Physics engine, vehicle manager, AI system, behaviour trees, traffic, weather, path planning
src/core/ - 2
User interface
PyQt6 main windows, control panels, editors, analytics dashboard
src/ui/ - 3
Rendering
OpenGL render engine, scene, lighting, shaders, particles
src/ui/rendering/ - 4
Recording and replay
Session recording and replay system
src/core/recording_system.py
Specification.
- Vehicles
- 7 types: sedan, SUV, truck, sports car, bus, motorcycle, emergency vehicle
- Physics
- Vehicle dynamics with simplified Pacejka tyre slip model, suspension, collisions, road-surface friction
- AI behaviours
- SAE J3016 autonomy levels 0-5, behaviour trees, state machines, PyTorch neural networks
- Environment
- Rain, snow, fog, wind, day/night lighting, traffic lights, pedestrian crossings
- Tools
- Scenario, behaviour and map editors; recording and replay; challenge system
- Analytics
- Real-time dashboard, CSV/JSON export, performance profiler
- Interface
- PyQt6 with dockable panels and OpenGL 3D viewport
- Runs on
- PyQt6 desktop application with OpenGL viewport