IGVC Ground Vehicle – CSE Senior Design

Summary

Abstract - The general purpose of the project.
Background - Real word applications of this project. Superficial overview of the project's end goal.
Project Requirements - Constraints on the project (Imposed by the IGVC Competition).
System Overview - Brief Description of the different "layers" present in the project.
Results - Final result after 2 semesters of work
Future Work - Work needed to be finished by future teams inheriting the project.

Team Name

IGVC-002 (Path-planning Team)

Timeline

Fall 2023 – Spring 2024

Students

Alfredo Reina Corona
Matthew McClure
Emmanuel Chinwuba
Ahmed Almudhi
Joshua Ngo
Tanner Gillis

Sponsor

Dr. Mcmurrough

Abstract

We are creating an autonomous vehicle with the goal of competing in the Intelligent Ground Vehicle Competition (IGVC). The vehicle will have the capability to autonomously navigate a course, following GPS waypoints while avoiding physical obstacles and avoiding crossing out of bounds of the course. Our team is one of two teams currently working on this project. Our focus is on developing a pathfinding system, enabling wireless communication between the vehicle and the remote-control station, and if time permits, finishing the physical development of the vehicle. We will ensure that the vehicle can plan an optimal path around the course while safely avoiding the penalties for hitting obstacles, leaving the course, or driving unsafely. While following these guidelines, it is our goal to navigate the course as quickly as possible to score the most points.

Background

Autonomous ground vehicles are a very relevant topic in today’s engineering world. Automated braking, lane assist, and other similar features are becoming more widely used in modern cars, helping to improve public safety on our roads. Automated taxi services are becoming more widespread as well, including Arlington’s own RAPID self-driving shuttles. While our project is nowhere near as complicated as a fully self-driving road vehicle, we were excited to have a chance to work on a project in this field.

Project Requirements

Be able to autonomously navigate a semi-random, static outdoor course.
Meet the IGVC Competition size requirements.
Have an emergency stop button.
Have light to indicate what mode the vehicle is in.
Be able to maintain a minimum speed of 1mph.
Be able to carry a 25lb payload.
Must be able to accomplish waypoint navigation.
Must have enough torque to climb ramp of unspecified incline.
Must be able to identify flat obstacles through camera.
Path-finding system must continually find the safest path for the vehicle to travel.

System Overview

SENSOR LAYER:
This layer holds the vehicle’s sensors. These sensors gather information from the environment and
then send that information to the controller layer for processing. The camera gathers images of the
environment to detect obstacles via object recognition. The LiDAR creates a map of the environment.
The LiDAR data is crosschecked with the camera’s data to understand where every object is in relation to the vehicle. GPS gives us the exact location of the vehicle. This information is sent to the controller layer for path planning.

CONTROLLER LAYER:
The controller layer consists of the cube orange, mission planner, and out ROS2 pathfinding algorithm. Most of the information is received from the sensor layer, but some is sent through sort of PID which allows for a continual position and orientation estimation. This layer will use ROS2’s native path-planning algorithm in to traverse the environment. Once an adequately safe path is found, the controller layer will send a signal to the hardware layer for the vehicle to move.

HARDWARE LAYER:
This layer will execute the demanded control signals from the software. Additionally, this layer will be responsible for all low-level operations such as powering, circuit current regulations, and motors. This layer consists of all the wires, batteries, Arduino, and H-Bridges.

Results

The pathfinding system was completed early, allowing our group to continue the physical development of the vehicle. The circuit management, Arduino programming, and wheel development are very deep into development, but are not complete.