all projects
PPO · Gazebo

navbot_ppo

Mapless mobile-robot navigation: a PPO motion planner that drives a TurtleBot to a goal using only sparse LiDAR and the target pose — no SLAM, no map.

Result Open-source companion to arXiv:2405.16266 Stars ★ 53 Tags PPO · ROS · Gazebo · LiDAR · PyTorch
View on GitHub Paper · arXiv:2405.16266

A reinforcement-learning motion planner that learns to navigate from scratch. The agent’s state is a compact 16-D vector — downsampled LiDAR ranges plus the relative goal — and its action is continuous wheel velocity. There is no global map and no classical planner in the loop: the policy itself is the planner.

Highlights

  • Mapless. Navigation from sparse LiDAR + target pose only — robust to unseen layouts.
  • PPO, PyTorch. Trained end-to-end; the enhanced-PPO formulation from the paper.
  • Reproducible. Dockerized Gazebo GUI so the simulation comes up with one command.
  • Sequential goals. The demo shows the robot reaching a series of targets while weaving around obstacles in real time.

This repository is the practical, runnable counterpart to my publication on safe mobile-robot navigation with enhanced PPO.