Adversarial Robotics

The Workshop on Adversarial Robotics will bring together researchers from the interdisciplinary areas of robotics, security, and AI, with the goal of identifying and codifying the challenges in the emerging area of adversarial robotics - developing secure, resilient, and robust robots that operate in adversarial settings.

Robots are becoming prominent. However, robots may pose serious threats. Compromised robots could quickly become dangerous physical entities capable of causing substantive safety issues. Adversarial activities can result in incorrect robot perception, adaptation, and decision making, which could cause anything from mission failures to harm to humans and environments. It is clear that security, resilience and robustness of robots is critical, and is increasingly recognized one of the major challenges remaining for robots to operate in domains subject to these risks. We believe that now is the appropriate time to start discussion towards solving this challenge.

Topics relevant to the agenda range from development of robot perception, adaptation, reasoning, and planning that are resilient to adversaries, application of robots operating in adversarial environments, cybersecurity design to secure robots, and theoretical and experimental analysis of attack vectors to robots. We will include speakers from beyond the typical robotics community, to identify connections with work in security and AI, and open up exciting avenues of research in adversarial robotics. This is the first workshop of its kind at an academic conference, and is intended to serve as a timely call to arm the academic community in response to the growing promise of this emerging field.

We invite submission of poster abstracts on topics within the theme of the workshop: Adversarial Robotics. We accept published and ongoing work. Authors of accepted abstracts will present their posters at the workshop. In addition, the pdf file of the posters along with their abstracts will be made available for general public on the workshop website. Topics of interest include but are not limited to:

A 60-minute poster session will be conducted that will allow authors of selected submissions to present work in this research area, and interact with other researchers and government/industry participants. Selected submissions* that are highly relevant to the workshop theme will be invited to give short (15-minute) oral spotlight presentations.

Time Speaker/Event Topic
13:25 - 13:30 Workshop Organizers Welcome and Introduction
13:30 - 14:00 Katia Sycara (CMU) Solution Space of Adversarial Robotics
14:00 - 14:30 Chad Tossell, Nathan Tenhundfeld (USAFA) A Balanced View of Problem and Solution Spaces
14:30 - 15:00 Coffee Break / Poster Session
15:00 - 15:30 Poster Session
15:30 - 16:00 Timothy Chung (DARPA) Problem Space of Adversarial Robotics
16:00 - 16:30 Nikolay A. Atanasov (UCSD) Solution Space of Adversarial Robotics
16:30 - 16:45 Lerrel Pinto (CMU) Spotlight Talk 1: Robust Adversarial Reinforcement Learning
16:45 - 17:00 Yulin Yang (UD) Spotlight Talk 2: Attack-Resilient Map-based Localization
17:00 - 17:30 Open Panel Discussion

Hao Zhang
Department of Computer Science
Colorado School of Mines

Christopher Reardon
U.S. Army Research Laboratory

Kun Sun
Department of Information Sciences and Technology
George Mason University

Chuan Yue
Department of Computer Science
Colorado School of Mines

Advisory Board

RSS 2018 Workshop on Adversarial Robotics