RADIUS: Risk-Aware, real-time trajectory Design In Uncertain Scenarios

RADIUS is a risk-aware real-time trajectory planning framework for autonomous driving. RADIUS uses Offline zonotope-based reachability analysis on the full order vheicle dynamics to compute the corresponding control-parametrized, over-approximative Forward Reachable Sets (FRS). Real-time trajectory planning is achieved by solving an optimization framework in real-time with the pre-computed, control-parametrized FRS being used to ensure vehicle safety up to a given threshold. The link to the project website can be found here.

Authors: Jinsun Liu* (jinsunl@umich.edu), Challen Enninful Adu* (enninful@umich.edu), Lucas Lymburner (llymburn@umich.edu), Vishrut Kaushik (vishrutk@umich.edu), Lena Trang (ltrang@umich.edu) and Ram Vasudevan (ramv@umich.edu).

*Equal Contribution

All authors are affiliated with the Robotics department and the department of Mechanical Engineering of the University of Michigan, 2505 Hayward Street, Ann Arbor, Michigan, USA.

Installation Requirements

RADIUS is built on Ubuntu 20.04 with ROS Noetic Distribution, and the algorithms are implemented in MATLAB and C++17.

Overview

0. Installation

Clone the RADIUS git repository and run the following from the top level:

./download-dependencies.sh

Run install.m.
In split.m, replace line 20 with cd(your_matlab_directory/toolbox/matlab/strfun) and line 22 with cd('your_CORA2018_directory/global functions/globOptimization'). Notice CORA2018 is downloaded automatically by ./download-dependencies.sh, so it should appear inside util.

Continue installation and setup by following instructions in the README for Full_Size_Vehicle_Simulation.

1. Offline reachability analysis

RADIUS adopts offline reachability analysis from REFINE. See FRS_generation for details.

2. Simulation