SHPE Omnidirectional Control Workspace - OmniCatcher

Introduction and Scope

This repository contains a ROS2 Humble workspace allowing for real time deployment of different pre-trained models for object capture. Students and other contributors can create their own controllers to test and deploy on the robot. Development is underway for human following and more.

Startup

First go to Jetson Containers and setup your Jetson to create the container using the command below:

git clone --recursive --depth=1 https://github.com/dusty-nv/jetson-inference &&
cd jetson-inference &&
docker/run.sh --ros=humble

Then, clone this repo into your workspaces directory:

git clone https://github.com/gabearod2/omni_control_ws

Then, run the container and build:

docker run -it --rm -v /home/shpe/workspaces:/workspaces omni_control_container bash -c "cd /workspaces && ls"
cd omni_control_ws
colcon build

Then, run the launch file for the controller based on the bounding box messages.

ros2 launch omni_launch omni_control.launch.py

Check out Embry-Riddle's SHPE chapter here.