Assessment and assurance of autonomous vehicle safety
Cyriac, James
Loading…
Permalink
https://hdl.handle.net/2142/109609
Description
Title
Assessment and assurance of autonomous vehicle safety
Author(s)
Cyriac, James
Issue Date
2020-12-02
Director of Research (if dissertation) or Advisor (if thesis)
Iyer, Ravishankar
Department of Study
Electrical & Computer Eng
Discipline
Electrical & Computer Engr
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
M.S.
Degree Level
Thesis
Keyword(s)
Autonomous Vehicle
Safety assessment
Fault Injection
Abstract
Safety is the biggest roadblock in deploying autonomous vehicles (AVs) on public roads as exemplified by recent headline-making accidents. Ensuring AV safety is as important as developing technology that enables self-driving functionality and requires more attention. This thesis has a few goals. First, it organizes and analyzes the current state-of-the-art methods for evaluating AV safety, elaborating on their advantages, drawbacks and potential directions for improvement. Second, it studies the recent Boeing 737 MAX crashes to highlight the importance of prioritizing the assessment and assurance of safety-critical systems like AVs over business interests. Then it demonstrates two methods of assessing AV safety: post-mortem data analysis and end-to-end fault injection on AVs. Post-mortem data analysis of the disengagement and crash reports from various manufacturers testing their AVs in California, USA, shows that AV technology needs further development to be as safe as human drivers and is not ready to be deployed unsupervised on public roads. The end-to-end fault injection framework can be used to assess and uncover safety-critical faults of the entire AV stack using a realistic driving simulator framework.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
