Spring 2016 Seminars
Design of Robotics and Embedded systems, Analysis, and Modeling Seminar (DREAMS)
The Design of Robotics and Embedded systems, Analysis, and Modeling Seminar (DREAMS) occurs weekly on Mondays from 4.10-5.00 p.m. in 250 Sutardja Dai Hall.
In Spring 2016, DREAMS joined forces with the Control Theory Seminar and the CITRIS People and Robots Seminar (CPAR).
The videos of the talks are posted on CITRIS CPAR Spring YouTube channel.
Information on the seminar series might be useful for potential speakers. If you have any questions about DREAMS, please contact Dorsa Sadigh. If you want to subscribe to our mailing list, please drop me a line.
Seminars from previous semesters can be found here.
Verifying Hybrid Systems
Jan 11, 2016, 4-5pm, 250 SDH, Ashish Tiwari, SRI International.
Most of the existing approaches for verification of hybrid systems are lifted versions of verification approaches that have worked well for discrete systems.
In this talk, we shall explore a few new verification techniques that are designed specifically for verifying hybrid systems. Specifically, we present abstraction techniques that abstract the dynamics, and not the state space, of the system. We also consider the option of abstracting the initial states and property, but not the state space and dynamics of the system. We will conclude by briefly discussing the larger landscape of techniques that are currently used for verifying hybrid systems.
Ashish Tiwari is a Senior Computer Scientist in the formal methods group of the Computer Science Laboratory at SRI International. He received his B.Tech and Ph.D. degrees in Computer Science from the Indian Institute of Technology, Kanpur and the State University of New York at Stony Brook in 1995 and 2000, respectively. His research interests lie in the areas of static program analysis, formal methods, automated deduction, automated synthesis, and symbolic computation. He is especially interested in using symbolic techniques and machine learning techniques in analyzing models of cyber-physical systems and models coming from various other domains, including systems biology.
Body Languages for physical Human-Robot Interaction
Jan 25, 2016, 3-4pm, 250 SDH, Antonio Bicchi, University of Pisa.
Modern approaches to the design of robots with increasing amounts of embodied intelligence affect human-robot interaction paradigms. The physical structure of robots is evolving from traditional rigid, heavy industrial machines into soft bodies exhibiting new levels of versatility, adaptability, safety, elasticity, dynamism and energy efficiency. New challenges and opportunities arise for the control of soft robots: for instance, carefully planning for collision avoidance may no longer be a dominating concern, being on the contrary physical interaction with the environment not only allowed, but even desirable to solve complex tasks. To address these challenges, it is often useful to look at how humans use their own bodies in similar tasks, and even in some cases have a direct dialogue between the natural and artificial counterparts.
Antonio Bicchi is Professor of Robotics at the University of Pisa, and Senior Scientist at the Italian Institute of Technology in Genoa. He graduated from the University of Bologna in 1988 and was a postdoc scholar at M.I.T. Artificial Intelligence lab in 1988Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½1990. He teaches Control Systems and Robotics in the Department of Information Engineering (DII) of the University of Pisa, leads the Robotics group at the Research Center "E. Piaggio'' of the University of Pisa since 1990, where he was Director from 2003 to 2012. He is an Adjunct Professor at the School of Biological and Health Systems Engineering of Arizona State University since 2013.
His main research interests are in Robotics, Haptics, and Control Systems in general. He has published more than 400 papers on international journals, books, and refereed conferences. He is Editor-in-Chief of the IEEE Robotics and Automation Letters, which he started in 2015. He is Program Chair of the IEEE Int.. Conf. Robotics and Automation (ICRA'16), has co-organized and chaired the first WorldHaptics Conference (2005), and Hybrid Systems: Computation and Control (2007). He served as the President of the Italian Association or Researchers in Automatic Control (2012-2013), as Editor in Chief of the Conference Editorial Board for the IEEE Robotics and Automation Society (RAS), as Vice President for Publications (2013-2014), for Membership (2006-2007), and as Distinguished Lecturer (2004-2006) of IEEE RAS. He is the recipient of several awards and honors. In 2012, he was awarded with an individual Advanced Grant from the European Research Council for his research on human and robot hands. Antonio Bicchi is a Fellow of IEEE since 2005.
Probabilistic Inference by Hashing and Optimization
Jan 25, 2016, 4-5pm, 250 SDH, Stefano Ermon, Stanford University.
Statistical inference in high-dimensional probabilistic models (i.e., with many variables) is one of the central problems of statistical machine learning and stochastic decision making. To date, only a handful of distinct methods have been developed, most notably (MCMC) sampling, decomposition, and variational methods. In this talk, I will introduce a fundamentally new approach based on random projections and combinatorial optimization. Our approach provides provable guarantees on accuracy, and outperforms traditional methods in a range of domains, in particular those involving combinations of probabilistic and causal dependencies (such as those coming from physical laws) among the variables. This allows for a tighter integration between inductive and deductive reasoning, and offers a range of new modeling opportunities.
Stefano Ermon is currently an Assistant Professor in the Department of Computer Science at Stanford University, where he is affiliated with the Artificial Intelligence Laboratory. He completed his PhD in computer science at Cornell in 2015. His research interests include techniques for scalable and accurate inference in graphical models, statistical modeling of data, large-scale combinatorial optimization, and robust decision making under uncertainty, and is motivated by a range of applications, in particular ones in the emerging field of computational sustainability. Stefano has won several awards, including two Best Student Paper Awards, one Runner-Up Prize, and a McMullen Fellowship.
Haptic perception, grasping and manipulation with simple robotic grippers
Feb 01, 2016, 4-5pm, 250 SDH, Matthew Mason, Carnegie Mellon University.
For about fifty years robotics researchers have been designing and testing robot grippers and hands. The designs vary dramatically in complexity, from a simple pair of tongs to hands with complexity approaching the human hand in some respects. The "Simple Hands" project at Carnegie Mellon seeks to demonstrate advanced manipulation capabilities with very simple hands, for example a gripper with a single motor and just a few sensors. With this system we have demonstrated grasping of objects, haptic recognition, pose estimation, change of grasp pose, and placing. Our approach uses physics models based on Newtonian mechanics and Coulomb friction, combined with machine learning techniques.
Matthew T. Mason earned the BS, MS, and PhD degrees in Computer Science and Artificial Intelligence at MIT, finishing his PhD in 1982. Since that time he has been on the faculty at Carnegie Mellon University. He was Director of the Robotics Institute from 2004 to 2014, and is presently Professor of Robotics and Computer Science. His prior work includes force control, automated assembly planning, mechanics of pushing and grasping, automated parts orienting and feeding, and mobile robotics. He is co-author of "Robot Hands and the Mechanics of Manipulation" (MIT Press 1985), co-editor of "Robot Motion: Planning and Control" (MIT Press 1982), and author of "Mechanics of Robotic Manipulation" (MIT Press 2001). He is a Fellow of the AAAI, and a Fellow of the IEEE. He is a winner of the System Development Foundation Prize and the IEEE Robotics and Automation Society's Pioneer Award.
Stratification, target set reachability and incremental enlargement principle
Feb 08, 2016, 4-5pm, 250 SDH, Lotfi A. Zadeh, University of California, Berkeley.
This paper presents a brief exposition of a version the concept of stratification, call it CST for short. In our approach to stratification, CST is a computational system in which the objects of computation are strata of data. Usually, the strata are nested or stacked with nested strata centering on a target set, T. CST has a potential for significant applications in planning, robotics, optimal control, pursuit, multiobjective optimization, exploration, search and other fields. Very simple, familiar examples of stratification are dictionaries, directories and catalogues. A multi-layer perceptron may be viewed as a system with a stratified structure. In spirit, CST has similarity to dynamic programing (DP), but it is much easier to understand and much easier to implement. An interesting question which relates to neuroscience is: Does the human brain employ stratification to store information? It would be natural to represent a concept such as chair, as a collection of strata with one or more strata representing a type of chair.
Underlining our approach is a model, call it FSM. FSM is a discrete-time, discrete-state dynamical system which has a finite number of states. The importance of FSM as a model derives from the fact that through the use of granulation and/or quantization almost any kind of system can be approximated to by a finite state system. A concept which plays an important role in our approach is that of target set reachability. Reachability involves moving (transitioning) FSM from a state w to a state in target state, T, in a minimum number of steps. To this end, the state space, W, is stratified through the use of what is refer as the incremental enlargement principle. It should also be noted that the concept reachability is related to the concept of accessibility in modal logic.
LOTFI A. ZADEH is a Professor in the Graduate School, Computer Science Division, Department of EECS, University of California, Berkeley. In addition, he is serving as the Director of BISC (Berkeley Initiative in Soft Computing).
Lotfi Zadeh is an alumnus of the University of Tehran, MIT and Columbia University. From 1950 to 1959, Lotfi Zadeh was a member of the Department of Electrical Engineering, Columbia University. He joined the Department of Electrical Engineering at UC Berkeley in 1959 and served as its Chair from 1963 to 1968. During his tenure as Chair, he played a key role in changing the name of the Department from EE to EECS.
Lotfi Zadeh held visiting appointments at the Institute for Advanced Study, Princeton, NJ; MIT, Cambridge, MA; IBM Research Laboratory, San Jose, CA; AI Center, SRI International, Menlo Park, CA; and the Center for the Study of Language and Information, Stanford University.
Lotfi Zadeh is a Fellow of the IEEE, AAAS, ACM, AAAI, and IFSA. He is a member of the National Academy of Engineering and a Foreign Member of the Finnish Academy of Sciences, the Polish Academy of Sciences, Korean Academy of Science & Technology, Bulgarian Academy of Sciences, the International Academy of Systems Studies, Moscow, and the Azerbaijan National Academy of Sciences. He is a recipient of the IEEE Education Medal, the IEEE Richard W. Hamming Medal, the IEEE Medal of Honor, the ASME Rufus Oldenburger Medal, the B. Bolzano Medal of the Czech Academy of Sciences, the Kampe de Feriet Medal, the AACC Richard E. Bellman Control Heritage Award, the Grigore Moisil Prize, the Honda Prize, the Okawa Prize, the AIM Information Science Award, the IEEE-SMC J. P. Wohl Career Achievement Award, the SOFT Scientific Contribution Memorial Award of the Japan Society for Fuzzy Theory, the IEEE Millennium Medal, the ACM 2001 Allen Newell Award, the Norbert Wiener Award of the IEEE Systems, Man and Cybernetics Society, Civitate Honoris Causa by Budapest Tech (BT) Polytechnical Institution, Budapest, Hungary, the V. Kaufmann Prize, International Association for Fuzzy-Set Management and Economy (SIGEF), the Nicolaus Copernicus Medal of the Polish Academy of Sciences, the J. Keith Brimacombe IPMM Award, the Silicon Valley Engineering Hall of Fame, the Heinz Nixdorf MuseumsForum Wall of Fame, the Egleston Medal, the Franklin Institute Medal, the Medal of the Foundation by the Trust of the Foundation for the Advancement of Soft Computing, the High State Award Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Friendship OrderÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½, from the President of the Republic of Azerbaijan, the Transdisciplinary Award and Medal of the Society for Design and Process Sciences, other awards and twenty-four honorary doctorates. He has published extensively (over 200 single-authored papers) on a wide variety of subjects relating to the conception, design and analysis of information/intelligent systems, and is serving on the editorial boards of over seventy journals.
Prior to the publication of his first paper on fuzzy sets in 1965, Lotfi ZadehÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½s work was concerned in the main with systems analysis, decision analysis and information systems. His current research is focused on fuzzy logic, semantics of natural languages, computational theory of perceptions, computing with words, extended fuzzy logic and Z-numbers.
Formal Methods for Highly Automated Driving
Feb 22, 2016, 4-5pm, 250 SDH, Werner Damm, Carl von Ossietzky UniversitÃ�ï¿½Ã¯Â¿Â½Ã�Â¯Ã�Â¿Ã�Â½Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¤t Oldenburg.
We discuss three complementary approaches jointly contributing for building safety cases for highly automated driving.
(1) Traffic Sequence Charts (TSCs), derived from Live Sequence Charts, offer a formal specification method for requirement capturing and scenario description for highly automated driving. They form the basis for adressing the following industrial needs 1. A formally defined compositional approach to generate all possible traffic environment situations from a parameterized set of atomic scenarios 2. A requirement analysis method for cooperative highly automated driving supported with methods for simulation based methods for completeness analysis and methods for proving consistency of requirements. 3. Specify conditions on the health state of the vehicle or the environmental conditions around the vehicle under which the specified service for highly automated driving is available 4. Supported by methods for virtual model-in-the loop testing, hardware-in-the loop testing, and vehicle testing of highly automated driving functions 5. Supported by methods for automatic generation of monitors for on-line supervision of assumptions and services for highly automated driving (2) Remorse free Strategies and optimality of world models We introduce the concept of remorse-free dominant strategies which allow to compare strategies for applications where winning strategies dont exist (such as for highly autonomous driving), using the intuitive concept of remorse. We call strategies Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½remorse-free dominantÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ if no other strategy could have done better in comparable environment situations, even if we add more observations about real-world artefacts to the world model. We can effectively test whether a model allows for remorse free dominant strategies after using methods such as predicate abstraction to reduce models to finite state models Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ thus such world models are optimal Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½adding more observations does not improve the strategic capabilities. We can effectively synthese remorse-free dominant strategies in optimal world models. We can effectively compute assumptions on the environment under which such strategies are in fact winning strategies.
(3) We present a virtual testbed for Human-in-the-Loop analysis of advanced driver assistance system, which allows for co-simulation and statistical model-checking of executable driver models based on empirically validated cogntive driver models, models of ADAS, models of vehicle dynamics and models of the traffic environment, developed jointly with the DLR Institute of Transportation, supporting TSCs
Werner Damm holds a Diploma in Computer Science and Mathematics from the University of Bonn (1976), and a PhD in Computer Science from the RWTH Aachen (1981, Best Dissertation Award) in formal semantics. In 1986 he received the Venia Legendi from the RWTH Aachen for his research in Computer Architecture. Since 1987 he is a full professor at the Carl von Ossietzky UniversitÃ�ï¿½Ã¯Â¿Â½Ã�Â¯Ã�Â¿Ã�Â½Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¤t Oldenburg, holding first the Chair for Computer Architecture, and since 2002, the Chair for Safety Critical Embedded Systems. He was a visiting professor at the Weizmann Institute of Sciences (1997, cooperating with A. Pnueli and D. Harel) and Uppsala University (2001, cooperating with B.Jonsson). He is the Scientific Director of the Transregional Collaborative Research Center AVACS (SFB/TR 14 Automatic Verification and Analysis of Complex Systems, www.avacs.org), and the Director of the Interdisciplinary Research Center on Critical Systems Engineering for Socio-Technical Systems (www.uni-oldenburg.de/en/cse/) . He is a member of acatech, the German National Academy of Science and Engineering.
He is a member of the Editorial Board of the Journal of Formal Methods in System Design, a member of the Steering Board of the CyberÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Physical Systems Week, and the Chairman of the Steering Board of the conference series on Hybrid Systems HSCC.
He is driving applied research in his roles as a member of the Board of Directors of the applied research institute OFFIS (www.offis.de), a not-for-profit association providing IT solutions for energy, health, and transportation, where he is responsible for its cross-sectorial research strategy and chairing the OFFIS R&D Division on Transportation.
He is the Chairman of the SafeTRANS association (www.safetrans-de.org), a not-for profit association with institutional membership, integrating leading companies and research institutes in the transportation domain in joint strategic projects on enhancing safety in transportation through IT-based solutions. He has been chairing the committee for the German National Roadmap for Embedded Systems and contributed in other roadmapping activities such as the acatech agenda CPS, the roadmap for industry 4.0 coordinated by the Forschungsunion, the roadmap Automotive Embedded Systems 2030 published jointly by SafeTRANS and VDA, and is currently chairing a roadmap process for Safety, Testing and Development Processes for Highly Automated Systems.
He is a co-founder and Chairman of the steering board of the European Institute for Complex Safety Critical Systems Engineering (www.eicose.eu), integrating large enterprises, SMEs and research organizations developing critical systems for aerospace, automotive, automation, rail and health applications. EICOSE has been driving the R&D strategy of the Joint Undertaking Artemis (www.artemis.org) in the areas of safety critical systems, human centered design, and hardware architectures through roadmapping and incubation of large projects implementing its roadmap. EICOSE has created a European wide innovation eco-system around the Artemis Tool Platform for Critical Systems Engineering, and has been recognized as a center of innovation excellence by the Artemis Industrial Association representing the private sector in the Joint Undertaking Artemis.
Werner Damm is a coÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½founder and member of the Board of BTC Embedded Systems AG (www.btcÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½es.org), the supplier for IBM RationalÃ�ï¿½Ã¯Â¿Â½Ã�Â¯Ã�Â¿Ã�Â½Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â´s Testing Solutions, and the provider of ISO 26262 compliant testing solutions for embedded automotive applications.
Embracing Complexity: A Fractal Calculus Approach to the Modeling and Optimization of Cyber-Physical Systems
Feb 29, 2016, 3-4pm, 250 SDH, Paul Bogdan, University of Southern California.
Cyber-physical systems (CPS) constitute a new generation of networked embedded systems that interweave computation, communication and control to facilitate our interaction with the physical world. They will stand among other application domains at the foundation of novel smart healthcare systems, which monitor individual physiological process across time and enable accurate disease prediction and health assessment. However, existing approaches to their modeling and optimization ignore important mathematical characteristics (e.g., non-stationarity, fractality). To face these challenges, we embrace the complexity of biological systems: instead of skirting around their non-linear variability. We propose a statistical physics inspired approach to CPS by encapsulating the observed mathematical characteristics of cyber and physical processes via a dynamical master equation. The first part of the talk is dedicated to explaining the benefits of this new approach, which facilitates a more accurate state-space modeling of Networks-on-Chip workloads, contributes to power savings and opens new possibilities for the dynamic optimization of large-scale systems. The second part focuses on a concrete example of a mathematical model based on fractional calculus concepts, which takes into account the dynamics of blood glucose characteristics (e.g., time dependent fractal behavior) and can be used to design an artificial pancreas that regulates insulin injection. Finally, the benefits of this mathematical framework will be also demonstrated in the context of interdependent networks by elucidating the brain-muscle interdependency with applications to brain-machine-body-interfaces and the brain functional connectivity.
Paul Bogdan is an Assistant Professor in the Ming Hsieh Department of Electrical Engineering at University of Southern California. He received his Ph.D. degree in Electrical and Computer Engineering from Carnegie Mellon University. His work has been recognized with a number of distinctions, including the 2015 NSF CAREER Award, the 2012 A.G. Jordan Award from the Electrical and Computer Engineering Department, Carnegie Mellon University for outstanding Ph.D. thesis and service, the 2012 Best Paper Award from the Networks-on-Chip Symposium (NOCS), the 2012 D.O. Pederson Best Paper Award from IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, the 2012 Best Paper Award from the International Conference on Hardware/Software Codesign and System Synthesis (CODES+ISSS), the 2013 Best Paper Award from the 18th Asia and South Pacific Design Automation Conference, and the 2009 Roberto Rocca Ph.D. Fellowship. His research interests include the theoretical foundations of cyber-physical systems, modeling and analysis of biological systems and swarms, understanding of neural and cognitive systems via new mathematical models, development of new control algorithms for dynamical systems exhibiting multi-fractal characteristics, modeling biological / molecular communication, development of fractal mean field games to model and analyze biological, social and technological system-of-systems, performance analysis and design methodologies for many core systems.
Computationally-Aware Cyber-Physical Systems
Mar 14, 2016, 4-5pm, 250 SDH, Jonathan Sprinkle, University of Arizona.
In this talk we describe hybrid model predictive controllers that switch between two predictor functions based on the uncontrollable divergence metric. The uncontrollable divergence metric relates the computational capabilities of the model predictive controller, to the error of the system due to model mismatch of the predictor function during computation of the model predictive control solution. The contribution of the work is its ability to trade off the accuracy of a predictive solution to the time values at which solutions will arrive. The results demonstrate the approach for control of ground vehicles as well as a vertical takeoff and landing aerial vehicle. The results are from joint work with Dr. Kun Zhang and Prof. Ricardo Sanfelice.
Jonathan Sprinkle is the Litton Industries John M. Leonis Distinguished Associate Professor of Electrical and Computer Engineering at the University of Arizona. In 2013 he received the NSF CAREER award, and in 2009, he received the UA's Ed and Joan Biggers Faculty Support Grant for work in autonomous systems. His work has an emphasis for industry impact, and he was recognized with the UA "Catapult Award" by Tech Launch Arizona in 2014, and in 2012 his team won the NSF I-Corps Best Team award. From 2003-2007 he was at the University of California, Berkeley, as a postdoctoral scholar and the Executive Director of CHESS. He graduated with the PhD and MS degrees from Vanderbilt University, and the BSEE degree from Tennessee Tech University.
The New Robotics Age: Meeting the Physical Interactivity Challenge
Mar 17, 2016, 4-5pm, 250 SDH, Oussama Khatib, Stanford University.
The generations of robots now being developed will increasingly touch people and their lives. They will explore, work, and interact with humans in their homes, workplaces, in new production systems, and in challenging field domains. The emerging robots will provide increased operational support in mining, underwater, and in hostile and dangerous environments. While full autonomy for the performance of advanced tasks in complex environments remains challenging, strategic intervention of a human will tremendously facilitate reliable real-time robot operations. Human-robot synergy benefits from combining the experience and cognitive abilities of the human with the strength, dependability, competence, reach, and endurance of robots. Moving beyond conventional teleoperation, the new paradigm Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ placing the human at the highest level of task abstraction Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ relies on robots with the requisite physical skills for advanced task behavior capabilities. Such connecting of humans to increasingly competent robots will fuel a wide range of new robotic applications in places where they have never gone before. This discussion focuses on robot design concepts, robot control architectures, and advanced task primitives and control strategies that bring human modeling and skill understanding to the development of safe, easy-to-use, and competent robotic systems. The presentation will highlight these developments in the context of a novel underwater robot, Ocean One, called O2, developed at Stanford in collaboration with Meka-Google Robotics, and KAUST.
Oussama Khatib received his PhD from SupÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Aero, Toulouse, France, in 1980. He is Professor of Computer Science at Stanford University. His research focuses on methodologies and technologies in human-centered robotics including humanoid control architectures, human motion synthesis, interactive dynamic simulation, haptics, and human-friendly robot design. He is a Fellow of IEEE. He is Co-Editor of the Springer Tracts in Advanced Robotics (STAR) series and the Springer Handbook of Robotics, which received the PROSE Award for Excellence in Physical Sciences & Mathematics. Professor Khatib is the President of the International Foundation of Robotics Research (IFRR). He has been the recipient of numerous awards, including the IEEE RAS Pioneer Award in Robotics and Automation, the IEEE RAS George Saridis Leadership Award in Robotics and Automation, the IEEE RAS Distinguished Service Award, and the Japan Robot Association (JARA) Award in Research and Development.
From Mimicry to Mastery: Creating Machines that Augment Human Skill
Mar 28, 2016, 4-5pm, 250 SDH, Gregory Hager, Johns Hopkins University.
We are entering an era where people will interact with smart machines to enhance the physical aspects of their lives, just as smart mobile devices have revolutionized how we access and use information. Robots already provide surgeons with physical enhancements that improve their ability to cure disease, we are seeing the first generation of robots that collaborate with humans to enhance productivity in manufacturing, and a new generation of startups are looking at ways to enhance our day to day existence through automated driving and delivery.
In this talk, I will use examples from surgery and manufacturing to frame some of the broad science, technology, and commercial trends that are converging to fuel progress on human-machine collaborative systems. I will describe how surgical robots can be used to observe surgeons Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½at workÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ and to define a Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½language of manipulationÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ from data, mirroring the statistical revolution in speech processing. With these models, it is possible to recognize, assess, and intelligently augment surgeonsÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ capabilities. Beyond surgery, new advances in perception, coupled with steadily declining costs and increasing capabilities of manipulation systems, have opened up new science and commercialization opportunities around manufacturing assistants that can be instructed Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½in-situ.Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ Finally, I will close with some thoughts on the broader challenges still be to surmounted before we are able to create true collaborative partners.
Gregory D. Hager is the Mandell Bellmore Professor of Computer Science at Johns Hopkins University. His research interests include collaborative and vision-based robotics, time-series analysis of image data, and medical applications of image analysis and robotics. He has published over 300 articles and books in these areas. Professor Hager is also Chair of the Computing Community Consortium, a board member of the Computing Research Association, and is currently a member of the governing board of the International Federation of Robotics Research. In 2014, he was awarded a Hans Fischer Fellowship in the Institute of Advanced Study of the Technical University of Munich where he also holds an appointment in Computer Science. He is a fellow of the IEEE for his contributions to Vision-Based Robotics, and has served on the editorial boards of IEEE TRO, IEEE PAMI, and IJCV. Professor Hager received his BA in Mathematics and Computer Science Summa Cum Laude at Luther College (1983), and his MS (1986) and PhD (1988) from the University of Pennsylvania. He was a Fulbright Fellow at the University of Karlsruhe, and was on the faculty of Yale University prior to joining Johns Hopkins. He is founding CEO of Clear Guide Medical.
The Syntax of action: A key to intelligent robots
Apr 18, 2016, 4-5pm, 250 SDH, Yiannis Aloimonos, University of Maryland, College Park.
Humanoid robots will need to learn the actions that humans perform. They will need to recognize these actions when they see them and they will need to perform these actions themselves. In this presentation, it is proposed that this learning task can be achieved using the manipulation grammar. Context-free grammars have been in fashion in linguistics because they provide a simple and precise mechanism for describing the methods by which phrases in some natural language are built from smaller blocks. Similarly, for manipulation actions, every complex activity is built from smaller blocks involving hands and their movements, as well as objects, tools and the monitoring of their state. Thus, interpreting a Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½seenÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ action is like understanding language, and executing an action from knowledge in memory is like producing language. Several experiments will be shown interpreting human actions in the arts and crafts or assembly domain, through a parsing of the visual input, on the basis of the manipulation grammar. This parsing, in order to be realized, requires a network of visual processes that attend to objects and tools, segment them and recognize them, track the moving objects and hands, and monitor the state of objects to calculate goal completion. These processes will also be explained in a new cognitive architecture called the Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Cognitive DialogueÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½ and we will conclude with demonstrations of robots learning how to perform tasks by watching videos of relevant human activities.
Yiannis Aloimonos is Professor of Computational Vision and Intelligence at the Department of Computer Science, University of Maryland, College Park, and the Director of the Computer Vision Laboratory at the Institute for Advanced Computer Studies (UMIACS). He is also affiliated with the Institute for Systems Research and the Neural and Cognitive Science Program. He was born in Sparta, Greece and studied Mathematics in Athens and Computer Science at the University of Rochester, NY (PhD 1990). He is interested in Active Perception and the modeling of vision as an active, dynamic process for real time robotic systems. For the past five years he has been working on bridging signals and symbols, specifically on the relationship of vision to reasoning, action and language.
Reach Control Problem
Apr 25, 2016, 4-5pm, 250 SDH, Mireille Broucke, University of Toronto.
We discuss a class of control problems for continuous time dynamical systems featuring synthesis of controllers to meet certain logic specifications. Such problems fall in the area of hybrid systems. Hybrid systems have been studied for some time; unfortunately the area has not delivered all that it promised: a comprehensive theory of control synthesis has remained elusive. Some work has been done at the high level on synthesis of controllers for logic specifications inspired by discrete event system theory. These approaches do not confront where the deeper challenge lies: a (hopefully) structural characterization of the intrinsic limits of a continuous time control system to achieve a non-equilibrium specification.
We study affine systems and logic specifications encoded as inequality constraints. Mathematically, the model is an affine system defined on a polytopic state space, and control synthesis typically yields piecewise affine controllers. By studying this special model, synthesis tools have been recoverable. The core synthesis problem has been distilled in the so-called Reach Control Problem (RCP). Roughly speaking, the problem is for an affine system xdot = A x + B u + a defined on a simplex to reach a pre-specified facet (boundary) of the simplex in finite time without first exiting the simplex. The significance of the problem stems from its capturing two essential requirements embedded in logic specifications: state constraints and trajectories reaching a goal set of states in finite-time.
In this talk I will give highlights of nearly 10 years of research on the RCP: solvability by affine feedback, continuous state feedback, time-varying affine feedback, and piecewise affine feedback; an associated Lyapunov theory; a geometric structure theory; and emerging applications.
Mireille Broucke obtained the BSEE degree in Electrical Engineering from the University of Texas at Austin in 1984 and the MSEE and PhD degrees from the University of California, Berkeley in 1987 and 2000, respectively. She was a postdoc in Mechanical Engineering at University of California, Berkeley during 2000-2001. She has six years of industry experience in control design for the automotive and aerospace industries. During 1993-1996 she was a program manager and researcher at Partners for Advanced Transportation and Highways (PATH) at University of California, Berkeley. Since 2001 she has been at the University of Toronto where she is a professor in Electrical and Computer Engineering. Her research interests are in hybrid systems, piecewise affine control, geometric control theory, and patterned linear systems.
Automated synthesis of control systems: A double abstraction scheme
May 02, 2016, 4-5pm, 250 SDH, Majid Zamani, Technical University of Munich.
Embedded control software plays a crucial role in many safety-critical applications: modern vehicles, for instance, use software to control steering, fuel injection, and airbag deployment. These applications are examples of cyber-physical systems (CPS), where software components interact tightly with physical systems. Although CPS have become ubiquitous in modern technology due to advances in computational devices, the development of core control software running on these systems is still ad hoc and error-prone. In this talk, I will propose a transformative design process, in which the controller code is automatically synthesized from higher-level correctness requirements. First, a compositional construction of abstractions of interconnected continuous systems is proposed. Those abstractions, themselves continuous systems, act as substitutes in the controller design process due to having possibly lower dimensions and simple interconnection topologies. Second, an automatic controller synthesis scheme is proposed by constructively deriving finite abstractions of infinite approximations of original continuous systems. The proposed automated synthesis of embedded control software holds the potential to develop complex yet reliable large-scale CPS while considerably reducing verification and validation costs.
Majid Zamani is an assistant professor in the Department of Electrical and Computer Engineering at Technical University of Munich where he leads the Hybrid Control Systems Group. He received a Ph.D. degree in Electrical Engineering and an MA degree in Mathematics both from University of California, Los Angeles in 2012, and an M.Sc. degree in Electrical Engineering from Sharif University of Technology in 2007. From September 2012 to December 2013, he was a postdoctoral researcher in the Delft Centre for Systems and Control at Delft University of Technology. Between December 2013 and May 2014, he was an assistant professor at Delft University of Technology.
Human machine teaming architectures: probabilistic shared control and the lower bounding property
May 04, 2016, 2-3pm, 240 Bechtel, Peter Trautman, Galois INC.
Shared control fuses operator inputs and autonomy inputs into a single command. However, if the environment or the operator exhibits Gaussian multimodality (common to search and rescue robots, teleoperated robots facing communication degradation, robotic prosthetics, and assistive driving technologies), many state of the art approaches (e.g., linear blending) are suboptimal with respect to safety, efficiency, and operator-autonomy agreement (see our recent paper at http://arxiv.org/pdf/1506.06784.pdf). To prove and rectify the above sub-optimalities, we introduced probabilistic shared control (PSC), which simultaneously optimizes autonomy objectives and operator-autonomy agreement under multimodal conditions. Importantly, because PSC optimizes operator-autonomy agreement, it exhibits stronger Ã�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½centaur" like properties than the state of the art: PSC naturally leverages the complementary abilities of the human and the machine.
More broadly, a disquieting paradox exists: human machine teams (in a general sense, to include both HCI and HRI) often perform worse than either the human or machine would perform alone. We thus argue for a strong but essential condition for any human machine team: team performance should never be worse than team member performance (the lower bounding property). Further, this property should be invariant to change in human capability, human modeling, autonomy capability, and environment. We consider case studies in which the lower bounding property is violated (linear blending under multimodal conditions, fully autonomous reactive path planners in human crowds) and where it is preserved (PSC using human models of varying fidelity, fully autonomous joint cooperative collision avoidance planners in human crowds).
Pete Trautman received his B.S. in Physics and Applied Mathematics from Baylor University in 2000. He then entered the United States Air Force, serving first as an analyst at the National Air and Space Intelligence Center, and then as a program manager/researcher at the Sensors Directorate (RYAT). In 2005, he returned to graduate school at Caltech, completing his Ph.D. in Control and Dynamical Systems in 2012. His thesis research focused on robot navigation in dense human crowds, the result of which was a probabilistic model of human robot cooperation and a 6 month case study in CaltechÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½s student cafeteria. From 2012 until early 2014, Pete was a Senior Engineer at the Boeing company, where he developed navigation and localization technology for commercial aircraft assembly robots. Additionally, he served as the sensing team lead for CaltechÃ�ï¿½Ã¯Â¿Â½Ã�ï¿½Ã�Â¢Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½Ã�ï¿½Ã�Â¯Ã�ï¿½Ã�Â¿Ã�ï¿½Ã�Â½s DARPA Grand Challenge entry in 2006, has consulted for Toyon Research and Applied Minds Inc., and has conducted research in physics at the Los Alamos National Laboratory, the University of California at San Diego, and Worcester Polytechnic Institute. He was a Best Paper Finalist at ICRA 2013 for his work on autonomous crowd navigation. He now works at Galois, Inc. in Portland Oregon.
The One Hundred Year Study on Artificial Intelligence: An Enduring Study on AI and its Influence on People and Society
May 10, 2016, 4-5pm, 250 SDH, Eric Horvitz, Microsoft Research, Redmond.
I will take the opportunity of a DREAM seminar to provide an update on the One Hundred Year Study on AI. I will describe the background and status of the project, including the roots of the effort in earlier experiences with the 2008-09 AAAI Panel on Long-Term AI Futures that culminated in the AAAI Asilomar meeting. I will reflect about several directions for investigation, highlighting opportunities for reflection and investment in proactive research, monitoring, and guidance. I look forward to comments and feedback from seminar attendees.
Eric Horvitz is a technical fellow and director of the Microsoft Research lab at Redmond. He has been elected fellow of AAAI, ACM, and the National Academy of Engineering (NAE). He served as president of the AAAI and has served on advisory boards for the Allen Institute for Artificial Intelligence, NSF, NIH, DARPA, and the Computing Community Consortium (CCC). More information can be found at http://research.microsoft.com/~horvitz.
|You are not logged in|
|©2002-2017 U.C. Regents|