Don’t Go That Way! Risk-Aware Decision Making for Autonomous Vehicles

Kasra Mokhtari, Kendra A. Lang, Alan R. Wagner

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Scopus citations


Accurately predicting risk or potentially risky situations is critical for the safe operation of autonomous systems. Risk refers to the expected likelihood of an undesirable outcome, such as a collision. We draw on an existing conceptualization of the risk to evaluate a robot’s options (e.g. choice of a path to travel). In this context, risk consists of two components: 1) the probability of an undesirable outcome computed by a Bayesian Network (BN) and 2) an estimate of the loss associated with the undesirable outcome. We demonstrate that our risk assessment tool is effective at computing the anticipated risk over a wide variety of the robot’s options and selecting the option with the lowest risk for two different types of autonomous systems: An Autonomous Vehicle (AV) operating near a college campus and a pair of Unmanned Aerial Vehicles (UAVs) flying from Washington DC to Baltimore. The method for assessing risk is used to identify higher risk routes, days to travel, and travel times for an autonomous vehicle and higher risk routes for a UAV.

Original languageEnglish (US)
Title of host publicationSocial Robotics - 12th International Conference, ICSR 2020, Proceedings
EditorsAlan R. Wagner, David Feil-Seifer, Kerstin S. Haring, Silvia Rossi, Thomas Williams, Hongsheng He, Shuzhi Sam Ge
PublisherSpringer Science and Business Media Deutschland GmbH
Number of pages12
ISBN (Print)9783030620554
StatePublished - 2020
Event12th International Conference on Social Robotics, ICSR 2020 - Golden, United States
Duration: Nov 14 2020Nov 18 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12483 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference12th International Conference on Social Robotics, ICSR 2020
Country/TerritoryUnited States

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • General Computer Science


Dive into the research topics of 'Don’t Go That Way! Risk-Aware Decision Making for Autonomous Vehicles'. Together they form a unique fingerprint.

Cite this