parkingLOT: Parking with Lidar Optimal Trajectory

Problem Statement

Parking is hard for human, so why don't we build a robot that does parking for us? In this project, we implement a parking system that ensures safety for both the vehicle, and the pedestrians around it.

Install:

Clone the repo recursively to a catkin workspace (named GEM-01 for our case):

cd GEM-01/src
git clone --recursive-submodules https://github.com/Safe-Autonomy/parkingLOT.git
cd parkingLOT/
pip install -v -e .

Build it!

cd .. # make sure to be in the src/ folder, not sure if needed
catkin build

Make sure to download the pre-trained YoloPv2 weights from their git repo into YOLOPv2/data/weights

Usage:

Real-world GEM Vehicle

Run these (on separate terminals, make sure to first source devel/setup.bash in each)

roslaunch basic_launch sensor_init.launch
roslaunch basic_launch visualization.launch

On another terminal, run parkingLOT:

roslaunch basic_launch lane_following.launch

Gazebo Simulation

TODO

Credit:

This project is the final project of group GEM-1 in ECE484: Principle of Safe Autonomy (Fall 2023). Developers are (of equal contribution): Zirui Wang, Anuj Sesha, Henry Che.