Undergrad Research Project - Path-Planning Parking Algorithms for Self-Driving Trailer Trucks

Spring 2017

Edward Ahn
Raj Rajkumar
Project description

This project will build a set of path planning parking algorithms for self-driving trailer trucks. Parking maneuvers for trailer trucks with physical constraints at loading docks can be challenging even for experienced human drivers. In contrast, while self-driving trucks can lead to better safety on both local roads and highways, these trucks must also able to park themselves in tight spots using accurate modeling of vehicle dynamics, kinematics and motion paths. This project will utilize the Gazebo 3-D graphics engine on ROS (Robot Operating System), and implement a realistic 3-D simulation environment where trailer trucks can successfully complete different kinds of parking maneuvers. This will result in a successful virtual simulation of trailer trucks parking themselves. A successful completion of this effort will lay the foundations for a follow-on project later that will implement these algorithms on a physical but scale-model of a trailer truck.

Flexible Deliverable Deadlines Jan 31 - Set up ROS/Gazebo Feb. 17 - Get a truck model working in Gazebo Mar. 9 - Be able to simulate movement of a truck, incorporating truck dynamics and kinematic models May 5 - Complete simulation of different parking maneuvers, such as tail-in, head-in, parallel parking and loading dock parking in constrained environments

Return to project list