rover_ros_ws

ROS packages for autonomous trajectory planning and tracking for IG32 Mecanum Ground Rover.

Setup

Built and tested on Intel NUC with Ubuntu 20.04 and ROS Noetic.

1. Install dependencies: pip install numpy scipy
2. Install pytorch: https://pytorch.org/get-started/locally/ (select Stable, Linux, Pip, Python, CPU)
3. Setup ROS with Arduino (https://maker.pro/arduino/tutorial/how-to-use-arduino-with-robot-operating-system-ros)
4. Flash Arduino code:
5. Install catkin tools: https://catkin-tools.readthedocs.io/en/latest/installing.html
6. Install ROS dependencies: sudo apt-get update and rosdep install --from-paths .
7. git clone https://github.com/adamdai/rover_ros_ws.git
8. cd rover_ros_ws
9. catkin build
10. source devel/setup.bash

Packages

controller

Nodes:

• traj_tracker.py: Listens for published nominal trajectories and handles state estimation and low-level control for tracking.

planner

Nodes:

• simple_planner.py: Publishes a single nominal trajectory parameterized by desired linear and angular velocity (can be provided as command-line arguments, i.e., rosrun planner simple_planner.py w_des v_des).
• sequence_planner.py: Publishes a pre-defined sequence of nominal trajectories
• nn_planner.py: Uses a pre-trained neural network to generate trajectories which avoid obstacles and reach a specified goal region.
• reachability_planner.py: Uses the neural network planner with an added reachability safety layer which checks the safety of planned trajectories and replans if unsafe.

sensing

Nodes:

• mocap.py: Interface with vrpn mocap topic to provide mocap measurements.

params

• params.py: Edit global parameters.

third_party

Contains all third party ROS packages used: vrpn_client_ros, cv_camera, apriltag,apriltag_ros

Usage

1. Start vrpn client: roslaunch vrpn_client_ros sample.launch
2. Start rosserial: rosrun rosserial_python serial_node.py _port:=/dev/ttyACM0
3. Start sensing nodes:
   • rosrun sensing mocap.py
   • rosrun cv_camera cv_camera_node
   • roslaunch apriltag_ros continuous_detection.launch
4. Start planning/control nodes:
   • rosrun controller traj_tracker.py
   • rosrun planner simple_planner.py or rosrun planner nn_planner.py or rosrun planner reachability_planner.py