JALAN-Sim Documentation

JALAN-Sim is a high-performance simulation library for autonomous ground vehicles (AGVs) designed for learning-based local navigation in complex environments. It provides efficient batch simulation capabilities with support for both CPU and GPU execution.

Overview

JALAN-Sim offers a comprehensive simulation framework that includes:

  • Multi-platform Support: Runs on both CPU and CUDA-enabled GPUs
  • Batch Simulations: Efficiently simulate thousands of agents simultaneously
  • Flexible Dynamics: Multiple vehicle models including bicycle, differential drive, and drift dynamics
  • Advanced Collision Detection: Circle and polygon-based collision models
  • Range Sensing: Bresenham and ray-marching algorithms for LiDAR simulation
  • Map Loading: Support for image-based occupancy grid maps
  • Python Bindings: Easy-to-use Python interface with NumPy integration

Use Cases

JALAN-Sim is ideal for:

  • Reinforcement Learning: Training navigation policies with massive parallel simulation
  • Path Planning Research: Testing algorithms across diverse environments
  • Robotics Education: Learning vehicle dynamics and control concepts
  • Algorithm Benchmarking: Comparing navigation approaches at scale

Getting Started

Ready to dive in? Check out our Installation Guide to get JALAN-Sim up and running on your system.

For detailed API documentation and examples, explore the other sections of this documentation.

License

JALAN-Sim is open source software. Please refer to the repository for current license information.