Integrated Adaptation and Model Inference for Safety Assurance of Cyber Physical Systems
DATE: Thursday, October 22, 2020
PLACE: 1819 Innovation Hub, University of Cincinnati (see below for directions)
- REGISTRATION & MASKS are REQUIRED
- This In-Person meeting is limited to non-students due to UC requirements.
- A separate registration is available for for virtually attendance for anyone, including students, for the presentation portion of the meeting- a Skype invitation will be sent to registered individuals.
- Health & Safety Protocols for Attendees. Masks are required, social distancing observed.
5:30 p.m.-5:45 p.m. Registration (Registration closes promptly at 5:45)
5:30 p.m. 6:00 p.m. Dinner & Social
6:00 p.m.- 6:30 p.m. Welcome address and Tour of Facility
6:30 p.m.- 8:30 p.m. Presentation
COST: $5 See information in Reservations
ABOUT THE MEETING: Cyber-Physical Systems (CPS) are systems in which physical and software components are very closely coupled, operate simultaneously on different spatial and temporal scales and interact with one another in ways not always entirely predictable at the time of design. Changes to any sub-component of a CPS while it is in operation can lead to unintended, but potentially dangerous, outcomes. The danger is particularly present when a Cyber Physical System employs reinforcement learning to adapt in response to the loss of efficacy. Reinforcement learning could adapt based on temporally and spatially local observations — but still be unready for conditions not seen during that adaptation.
This presentation will focus on a method to meld reinforcement learning and model inference in a self-adapting Flapping-Wing Micro Air Vehicle (FW-MAV). The salient features of the method are that it does not require system evaluation beyond that, one would use for adaptation and can that it can use “indirect” information that would be available during normal operation instead of dedicated, off-line, model verification. The ability to trust our increasingly complex systems is a critical need. Therefore, this presentation will also discuss the intended long-term development of these and related techniques to ensure ongoing verification and validation capability for other cyber-physical systems even in the presence of both intentional and unintentional injections of uncertainty.
ABOUT THE PRESENTER: Dr. John Gallagher received his B.S., M.S., and Ph.D. degrees in computer engineering from Case Western Reserve University. He is currently a Professor in the Department of Electrical Engineering and Computer Science at the University of Cincinnati. He conducted research in machine learning and A.I. as applied to robotics and control and analog VLSI neuromorphic computation and studies topics in machine learning and meta-heuristic search, verification, validation, and security of adaptive cyber-physical systems, neuromorphic computation, and adaptive autonomous robots.
MENU SELECTIONS: Buffet Menu: Box meal
LOCATION: 1819 Innovation Hub- University of Cincinnati
2900 Reading Road, Cincinnati, OH 45206
(Intersections of E. University Ave, Lincoln, and Reading at I-71)
(Lat Long) 39.132685, -84.497100
Entrance: Ground level at south entrance adjacent to the Children’s parking lot.
Recommended parking in the Children’s lot- free for this event.
RESERVATIONS: http://www.ieeecincinnati.org/meetings/. Please click on the appropriate link and complete the reservation. (Note: Meeting list on webpage is slow to load on some browsers)
(In Person) Cost is $5 for advanced registration. This In-Person meeting is limited to non-students due to UC requirements. Students and others may attend virtually and will receive connection information after registering- this is a separate registration. Reservations close at 11:59 PM on Wednesday October 21, 2020.
DINNER RESERVATION CANCELLATION POLICY
An email to firstname.lastname@example.org prior to the close of reservations is required to properly cancel your reservation.
All Reservations must be made by October 21st, 2020
WALK-INS (those without reservations): Walk-ins NOT AVAILABLE FOR THIS MEETING.
PE CREDITS: Depending on the subject matter, attendance at IEEE Cincinnati Section Meetings now qualifies the attendee for Professional Development Hours towards renewal of Professional Engineers Licenses. Required documentation will be available following the meeting if qualified! The Section Meetings also provide a great opportunity to network with fellow engineers in the area.