CPS Events
CPSRC Seminar Series: Making aerial robotics safer in the face of external disturbances
Abstract:
Flying robots, such as multicopters, are increasingly becoming part of our everyday lives, with current and future applications including personal transportation, delivery services, entertainment, and aerial sensing. These systems are expected to be safe and to have a high degree of autonomy. This talk will discuss the dynamics and control of multicopters, with a focus on making these vehicles more robust to external disturbances, and component failures. Finally, we will present the application of a failsafe algorithm to a fleet of drones performing as part of a live theater performance on New York’s Broadway.
Bio:
Mark W. Mueller joined the mechanical engineering department at UC Berkeley in September 2016. He completed his PhD studies, advised by Prof. Raffaello D’Andrea, at the Institute for Dynamic Systems and Control at the ETH Zurich at the end of 2015. He received a bachelors degree from the University of Pretoria, and a masters from the ETH Zurich in 2011, both in Mechanical Engineering.
Watch the seminar on our YouTube channel: https://youtu.be/aOznH9NUh2Y
CPSRC Seminar Series: Real-time Analytics and Scale-out Machine Learning with FPGA Key-Value Store
Abstract:
Key Value Store (KVS) provides a highly scalable means to store and retrieve distributed data over a network. In datacenters, high performance KVS allow large numbers of machines to share data by reading and writing key/value pairs over high-speed Ethernet. Algo-Logic has implemented a scaled-up KVS using Field Programmable Gate Array (FPGA) logic that achieved record-setting low latency, high throughput, and low power consumption. In this talk, it will be shown how this FPGA KVS was scaled out to accelerate machine learning for self-driving cars using a Markov Decision Process (MDP). Parallel systems were put together with the FPGA KVS to scale up machine learning and perform real-time decision making for 30 self-driving cars in a simulated highway driving environment.
Bio:
John W. Lockwood is an expert in building FPGA-accelerated applications. He is CEO of Algo-Logic Systems, Inc. and has founded three companies in the areas of low latency networking, Internet security, and electronic commerce. In industry, he worked at the National Center for Supercomputing Applications (NCSA), AT&T Bell Laboratories, IBM, and Science Applications International Corp (SAIC). In academia, he managed the NetFPGA program at Stanford University from 2007 to 2009 and grew the Beta program 10 to 1,021 cards deployed worldwide. As a tenured professor, he created and led the Reconfigurable Network Group within the Applied Research Laboratory at Washington University in St. Louis. He has published over 100 papers and patents on topics related to networking with FPGAs and served as served as principal investigator on dozens of federal and corporate grants. He holds BS, MS, PhD degrees in Electrical and Computer Engineering from the University of Illinois at Urbana/Champaign and is a member of IEEE, ACM, and Tau Beta Pi.
Watch the seminar on our YouTube channel:
Part 1: https://youtu.be/YNZK8V0r0uQ
Part 2: https://youtu.be/tF8Li59qUjg
CPSRC Seminar Series: Machine learning in oceanography: How algorithms and recent developments in underwater imaging will change the way we explore the ocean
Abstract:
The midwater region of the ocean (below the ocean surface and above the seafloor) is one of the largest ecosystems on our planet, yet remains one of the least explored. This region is home to processes and marine organisms we know almost nothing about, and necessarily links what’s happening in the atmosphere to the deepest depths of the ocean. Although significant advances in underwater vehicle technologies have improved access to midwater, methods for synthesizing this data are sorely needed as persistent observation platforms are utilized in the future. Here we present new imaging technologies (DeepPIV, an instrumentation package affixed to a remotely operated vehicle that quantifies fluid motions from the surface of the ocean down to 4000 m depths) and observational platforms (Mesobot, an autonomous underwater vehicle that uses stereo cameras to track underwater targets rated to 1000 m) that will enable investigations of the ocean’s midwaters in novel ways. Recently funded efforts to mine MBARI’s 30-year, expertly curated video database to generate an “ImageNet of the ocean” will also be presented. If successful, these efforts will lead to unprecedented observations of one of the least explored regions on our planet.
Bio:
Kakani received her PhD in Bioengineering at the California Institute of Technology and specializes in biological fluid mechanics and in situ imaging methods. She is currently a Principal Engineer and Principal Investigator at MBARI, with funding provided by the Packard Foundation and the National Science Foundation. Kakani has been named a National Geographic Emerging Explorer in 2011 and a Kavli Research Fellow in the National Academy of Sciences in 2013.
CITRIS/CPAR Control Theory and Automation Symposium | 1st NorCal Control Workshop
Symposium Theme:
Current challenges and future directions in control and automation.
CITRIS and the Banatao Institute, People and Robots Initiative (CPAR) Control Theory and Automation Symposium will be held on Friday, April 27, 2018, 10 am - 5 pm at UC Santa Cruz. This symposium will kick off the 1st NorCal Control Workshop, an annual event providing a forum to bring together students, postdocs and faculty from various universities, as well as representatives from industry, in the Northern California region working in the broad area of systems and control to share knowledge and build new connections.
This inaugural event is organized by CITRIS and the Banatao Institute, People and Robots Initiative (CPAR), and the Cyber-Physical Systems Research Center (CPSRC) at UC Santa Cruz and focuses on a timely theme to the field of systems and control. A goal of the symposium is to spark discussions leading to answers to the following questions: What are the key challenges in the development of control and automation solutions to the complex problems of today? What are unique future opportunities and problems where control and automation would play a key role? The event features two keynote talks, a panel with systems and control experts from academia and industry on current challenges and future directions, as well as a poster and networking session.
Event Program:
(with corresponding time code in video of proceedings on the CPSRC YouTube channel)
00:00 - 00:25 -- Welcome - Prof. Ricardo Sanfelice (UC Santa Cruz)
00:25 - 05:41 -- Introduction - Dean Alexander Wolfe (UC Santa Cruz)
05:58 - 20:40 Mengqiao Yu (UC Berkeley) - Making Intersections Safer with Intersection Intelligence Control System
20:49 - 35:00 -- Nathan Bucki (UC Berkeley) - Improved Quadcopter Disturbance Rejection using Added Angular Momentum
35:02 - 47:35 -- Erik Kiser (Naval Postgraduate School) - The Impact of Missions and Technologies on Contingency Base Fuel Consumption
47:42 - 01:02:50 -- Richard Shaffer (UC Santa Cruz) - Open-Loop Optimal Path Planning for a Nonlinear Flexible Double Gimbal with Parameter Uncertainty
01:02:53 - 01:13:44 -- Sina Dehghan (UC Merced) - PID2018 Benchmark Challenge: Model Predictive Control With Conditional Integral Control Using A General Purpose Optimal Control Problem Solver - RIOTS
01:13:59 - 02:12:56 -- Industry Keynote - Speaker: P.K. Menon (Optimal Synthesis Inc.) -- Title: Dynamics and Control of Air Traffic
02:13:15 - 02:14:35 -- Post-lunch Address Prof. Ken Goldberg (UC Berkeley)
02:14:36 - 02:30:12 -- Berk Altin (UC Santa Cruz) - Predictive Control of Hybrid Dynamical Systems
02:30:14 - 02:41:45 -- Gang Chen (UC Davis) - Formal Interpretation of Cyber-Physical System Performance with Temporal Logic
02:41:57 - 02:54:30 -- Pierre-Jean Meyer (UC Berkeley) - Sampled-data Reachability Analysis using Sensitivity and Mixed-monotonicity
02:54:47 - 03:11:35 -- Mo Chen (Stanford University) - A Differential Game Approach to Real-time Robust Planning
03:12:07 - 03:25:43 -- Sylvia Herbert (UC Berkeley) - Safe Control of Autonomous Dynamic Systems for Real-time Planning
03:25:45 - 04:14:00 -- Academia Keynote - Speaker: Prof. Arthur J. Krener (Naval Postgraduate School) - Title: Computational Issues in Nonlinear Control and Estimation
04:14:35 - 05:35:00 -- Panel discussion - Theme: Emerging Trends and Future Directions in Control Theory and Automation -- Panelists: Martin Sehr (Siemens), Murat Arcak (UC Berkeley), Stefano Carpin (UC Merced), Arthur J. Krener (NPS), Sanjay Lall (Stanford), P.K. Menon (Optimal Synthesis Inc.)
Watch the symposium on our YouTube channel: https://youtu.be/2AREYKF4pAE
| Attachment | Size |
|---|---|
 Slides from Prof. Arthur Krener's keynote address. | 340.18 KB |
| Example files from Prof Arthur Krener's keynote address. | 87.16 KB |
CPSRC Seminar Series: Indoor Human Information Acquisition from Physical Vibrations
Abstract:
The number of everyday smart devices (such as smart TV, Samsung SmartThings, Nest, Google Home, etc.) is projected to grow to the billions in the coming decade. The Cyber-Physical Systems or Internet of Things systems that consist of these devices are used to obtain human information for various smart building applications. Different sensing approaches have been explored, including vision-, sound-, RF-, mobile-, and load-based methods. The general problems faced by these existing technologies are their sensing requirements (e.g., line-of- sight, high deployment density, carrying a device) and intrusiveness (e.g., privacy concerns).
In this talk, I will introduce my research on non-intrusive indoor human information acquisition through ambient structural vibration, which I call ’structures as sensors’. People’s interaction with structures in the ambient environment (e.g., floor, table, door) induces those structures to vibrate. By capturing and analyzing the vibration response of structures, we can indirectly infer information about the people causing it. However, challenges remain. Due to the complexity of the physical world (both structures and people), sensing data distributions can change significantly under different sensing conditions. Therefore, accurate information learning through a data-driven approach requires a large amount of labeled data, which is costly and difficult if not impossible to obtain in sensing applications. My research addresses these challenges by utilizing physical insights to guide the sensing process. Specifically, my system can robustly learn human information from limited labeled data distributions by iteratively expanding the labeled dataset. With insights into the relationship between changes of sensing data distributions and measurable physical attributes, the expansion order is guided by measured physical attributes to ensure a high learning accuracy in each iteration.
Bio:
Shijia Pan received her Bachelor's degree in Computer Science and Technology from University of Science and Technology of China and will receive a Ph.D. degree in Electrical and Computer Engineering at Carnegie Mellon University in 2018. Her research interests include cyber-physical systems, Internet-of- Things (IoT), and ubiquitous computing. She worked in multiple disciplines and focused on indoor human sensing through ambient structural vibrations. She has published in both top-tier Computer Science ACM/IEEE conferences (IPSN, UbiComp) and high-impact Civil Engineering journals (Journal of Sound and Vibration, Frontiers Built Environment). She is the recipient of numerous awards and fellowships, including Nick G. Vlahakis Graduate Fellowship, Google Anita Borg Scholarship, Best Poster Awards (SenSys, IPSN), Best Demo Award (Ubicomp), Best Presentation Award (SenSys Doctoral Colloquium), and Audience Choice Award (BuildSys) from ACM/IEEE conferences.
Watch the seminar on our YouTube channel: https://youtu.be/1nAKUuvzUFc
