Mini-Workshops

Visible to the public 

Mini-Workshops

All Mini-Workshops will run concurrently from 1:15 to 3:00 p.m. on Tuesday, November 8.
The Mini-Workshops will be available to virtual participants via Zoom sessions. The Zoom links for the sessions will be posted on this page 15 minutes prior to the start of the session.


1. Artificial Intelligence (AI) for Healthcare
2. Ensuring Trustworthiness in AI-Enabled Cyber-Physical Systems
3. Intelligent Autonomous Cyber-Physical Systems: Design, Verification, and Certification
4. Socially Responsible Human Cyber-Physical Systems
5. Workshop on the Future of Cyber-Physical Systems

 

Artificial Intelligence (AI) for Healthcare
 

Organizers: Asimina Kiourti (The Ohio State University) and Fatemeh Afghah (Clemson University)

Artificial Intelligence (AI) is adding new capabilities to traditional engineering and computer science approaches used in the biomedical sector for sensor design, optimization, and data processing. In turn, unprecedented opportunities are brought forward for the future of medical diagnostics and treatment, promising personalized healthcare provision and a shift from reactive to proactive medicine. With a focus on healthcare, this workshop will discuss advances in AI algorithms and techniques, applications, and future opportunities.

Format and Speakers:

Two 20-minute presentations followed by 10-minute Q&A

  • Rance Cleaveland (University of Maryland) - “Data Mining vs. Data ‘Mine’-ing: AI/ML in Medicine”
  • Rose Faghih (New York University) - “Wearable Computing for Mental Well-Being”

45-minute interactive panel moderated by the workshop organizers. Panelists:

  • Rance Cleaveland (University of Maryland)
  • Rose Faghih (New York University)
  • Payam Heydari (University of California Irvine) 
  • Jia Liu (Harvard University)
  • John Stankovic (University of Virginia)

 

Ensuring Trustworthiness in AI-Enabled Cyber-Physical Systems
 

Organizer: Bruno Sinopoli (Washington University in St. Louis)

Trustworthy autonomy remains elusive, despite tremendous advances towards autonomous AI systems across a number of disciplines and a myriad of applications. A fundamental challenge shared by many autonomous systems is to ensure that their behavior is trustworthy in two distinct senses: social and technical. From a social perspective, trust is about how humans relate to and interact with autonomous systems, and its theoretical foundations remain poorly understood. Technical trustworthiness, in turn, relates to safety, security, and reliability of autonomous systems. While these are well-understood in traditional systems and control, they remain largely open problems when modern AI enters the control loop. This workshop will discuss challenges and research opportunities in this growing area of interest.

Speakers:

  • Ilya Kolmanovsky, Professor, Aerospace Engineering, University of Michigan, “Control of Safety Critical Systems under Uncertain Computations”
  • Andrew Clark, Associate Professor, Washington University in St Louis, “Safe and Resilient AI-CPS in Adversarial Environments”
  • Mahnoosh Alizadeh, Associate Professor, University of California, Santa Barbara, “Data-driven Safe Pricing in Safety Critical Infrastructures”
  • Aaron Ames, Professor, Caltech, “Enabling Trust with Safe CPS”

Panelists:

  • Huafeng Yu, Senior Researcher, Boeing Research and Technology
  • Sandip Roy, Professor, Washington State University
  • Dimitra Panagou, Associate Professor, University of Michigan

 

Intelligent Autonomous Cyber-Physical Systems: Design, Verification, and Certification


Organizers: Pierluigi Nuzzo (University of Southern California), Luca Carlone (Massachusetts Institute of Technology) and Peng Wei (The George Washington University)

Correctness, safety, and dependability assurance is of utmost importance in mission-critical cyber-physical systems. In these systems, as software is expected to perform increasingly more sophisticated tasks that were previously allocated to humans, one of the biggest challenges to full autonomy is arguably in showing that increasingly sophisticated artificial intelligence (AI) and autonomous software functions will still satisfy all the stringent correctness, safety, and dependability requirements. What could be a viable path toward accelerating the deployment of autonomy and AI in mission-critical systems while ensuring their exemplary safety and dependability records? What are the science and technology enablers that would accelerate the certification of AI-enabled mission-critical systems? This workshop will discuss a broad range of topics toward enhancing assurance of intelligent autonomous cyber-physical systems, including design-time and runtime verification and validation methods, correct-by-construction approaches, and state-of-the-art techniques to produce assurance evidence that can accelerate certification. A set of short presentations will be followed by a panel discussion moderated by the organizers.

Speakers and Panelists:

  • Francesco Borrelli, Professor, Mechanical Engineering, U.C. Berkeley, “Safe Learning in Autonomous Systems”

  • Samuel Coogan, Associate Professor, Electrical and Computer Engineering, Georgia Tech, “Fast Reachability as a Building Block for Verified Autonomy”

  • Sriram Sankaranarayanan, Professor, Computer Science, University of Colorado, Boulder, “Predictive Runtime Assurance for Autonomous Systems”

  • Hazem Torfah, Postdoctoral Researcher, U.C. Berkeley, “Learning Monitorable Operational Design Domains for Assured Autonomy”

  • Timothy E. Wang, Autonomous and Intelligent Systems, Raytheon Technologies Research Center, “Design and Assurance of Autonomy”

 


Socially Responsible Human Cyber-Physical Systems

Organizers: Meeko Oishi (University of New Mexico), Ufuk Topcu (UT Austin), and Tichakorn Wongpiromsarn (Iowa State University)

As autonomy becomes commonplace, human CPS must be designed not merely to be correct and functional, but more broadly, to be socially responsible. Socially responsible human CPS must meet objectives beyond those typical goals of safety and performance, so that, depending on the application, they are customizable, responsive, ethical, free of blame, and socially aware. Addressing such complex objectives will require the CPS community to draw upon interdisciplinary expertise in control theory, human-machine interaction, formal methods, ethics, autonomous systems, human factors, cognitive psychology, cognitive control, and other related areas. What novel paradigms in the design of computing, communication, control, and learning are needed to enable socially responsible human CPS? How can we integrate qualitative, or subjective, elements of social responsibility into the development of new theories, models, methods, and algorithms for human CPS? This workshop will engage attendees on a broad range of topics, including characterization of social responsibility, advances needed to enable socially responsible human CPS, and opportunities and resources necessary to create such advances.

Speakers and Panelists:

  • Anuradha Annaswamy, Senior Research Scientist, Mechanical Engineering, Massachusetts Institute of Technology

  • Lu Feng, Assistant Professor, Computer Science, University of Virginia

  • Todd Murphey, Professor, Mechanical Engineering, Northwestern University

  • Dorsa Sadigh, Assistant Professor, Computer Science, Stanford University

 

Workshop on the Future of Cyber-Physical Systems

Organizer: Simone Silvestri (University of Kentucky)

The National Science Foundation (NSF) initiated its Cyber-physical systems (CPS) program more than a decade ago. As a result, there have been many foundational solutions developed to address various research challenges in modeling, design, and control of trustworthy, real-time, and dependable cyber-physical platforms in different application domains. The proposed solutions in CPS technologies are transforming the way people interact with such engineered systems. While tremendous progress has been made in advancing CPS technologies, the demand for innovation across application domains is driving the need to accelerate research to keep pace. 

This workshop brings together experts from different CPS domains to discuss, identify, and debate challenges and opportunities of future CPS research. More specifically, the attendees will discuss i) important research challenges, ii) exciting opportunities for CPS research in the next few years, iii) lessons learned, including what worked and what didn’t work, and iv) ideas for research/technology transition to industry and society.

Panelists will provide a short presentation followed by a panel discussion.

Speakers and Panelists:

  • Panagiota  Karava, Purdue University
  • George J. Pappas, University of Pennsylvania
  • Miroslav Pajic, Duke University
  • Soumik Sarkar, Iowa State University
  • Saman Zonouz, Georgia Tech