Skip to content

GRAIC — Racing Simulation

Overview

Develop a planner and controller for [GRAIC](https://popgri.github.io/Race/), an international autonomous racing competition. The project progresses from using ground truth inputs to completing obstacle-laden tracks using vision only.
GRAIC racing simulation

Objectives

  • Complete the tracks using ground truth waypoints and obstacle information, beating the baseline scores with minimal collisions
  • Complete the tracks using vision only (no ground truth inputs)

System Modules

Module Description
Controller Follows ground truth waypoints around the track
Planner Plans around obstacles and integrates with the controller
Vision module Extracts waypoint and obstacle information from camera input

Milestones

# Time Milestone Description
1 First review Setup Setup simulator and become familiar with it.
2 Controller Develop a controller for ground truth waypoints
3 Planner + integration Develop an obstacle planner and integrate it with the controller
4 Vision module Develop the vision module to extract required track and obstacle information
5 Optimization Optimize the full system to minimize collisions and lap time
6 Final presentation Final race Final race at the end of the semester and launch project site

Bonus Points

  • Leaderboard performance: top-performing teams earn bonus points
  • Successfully completing the track using vision with obstacles

Primary Mentor

Hanna

References