Autonomous Robot Competition

Developing an autonomous robot in ROS Gazebo to navigate through an obstacle course and read clue boards using a trained character recognition model.

Getting Started

These instructions will set up the environment to deploy the simulation, score_tracker, and controller and image_processing nodes simulaneously.

Usage

A few examples of useful commands and/or tasks.

Build Environment:

cd ~/ros_ws
catkin_make

Source environment:

source ~/ros_ws/devel/setup.bash

Start the simulated world:

cd ~/ros_ws/src/2024_competition/enph353/enph353_utils/scripts
./run_sim.sh -vpg

Start score_tracker:

cd ~/ros_ws/src/2024_competition/enph353/enph353_utils/scripts
./score_tracker.py

To launch image processing and controller node:

cd ~/ros_ws/src/controller/launch
roslaunch my_launch.launch

These three will need to be launched in 3 separate terminals. The simulation must be launched first, followed by a score tracker and controller.

To view the simulation from the robot's camera feed:

rosrun rqt_image_view rqt_image_view

Competition Repository

The Competition repository and setup can be found here: https://github.com/stevenxu27/ENPH353_Competition_5

Branches

• main: Most up-to-date implemented drive and image recognition.
• image_processing: Optimizing different models to train the most efficient model for character recognition from a different perspective.
• imitation_learning_drive: Training our imitation learning drive around the course.
• initial_drive_testing: Basic code for time trials.