CSS Day 2022 CSS Day 2022 Presentations Welcome Message from the CSS-Day Task Force - Jing Sun Tech Showcase 1: Robot Ecology: Control Design for Long Duration Autonomy - Magnus Egerstedt Abstract: When robots are to be deployed over long time scales, optimality should take a backseat to “survivability”, i.e., it is more important that the robots do not break or completely deplete their energy sources than that they perform certain tasks as effectively as possible. For example, in the context of multi-agent robotics, we have a fairly good understanding of how to design coordinated control strategies for making teams of mobile robots achieve geometric objectives, such as assembling shapes or covering areas. But, what happens when these geometric objectives no longer matter all that much? In this talk, we consider this question of long duration autonomy for teams of robots that are deployed in an environment over a sustained period of time and that can be recruited to perform a number of different tasks in a distributed, safe, and provably correct manner. This development will involve the composition of multiple barrier certificates for encoding tasks and safety constraints through the development of non-smooth barrier functions, as well as a detour into ecology as a way of understanding how persistent environmental monitoring can be achieved by studying animals with low-energy life-styles, such as the three-toed sloth. Snake Robots - Kristin Y. Pettersen Abstract: Snake robots are motivated by the long, slender and flexible body of biological snakes, which allows them to move in virtually any environment on land and in water. Since the snake robot is essentially a manipulator arm that can move by itself, it has a number of interesting applications including firefighting applications and search and rescue operations. In water, the robot is a highly flexible and dexterous manipulator arm that can swim by itself like a sea snake. This highly flexible snake-like mechanism has excellent accessibility properties; it can for instance access virtually any location on subsea energy installations, move into the confined areas of shipwrecks, inside ice caves, or be used for observation of biological systems. Furthermore, not only can the swimming manipulator access narrow openings and confined areas, but it can also carry out highly complex manipulation tasks at this location since manipulation is an inherent capability of the system. Control and Learning for Legged Locomotion - Wei Zhang Abstract: Legged robots have attracted considerable attention from both academia and industry recently, owing to their superior ability of traversing complex terrains as compared with wheeled mobile robots. They are well suited for applications such as surveillance and monitoring in cluttered environments, search and rescue, last-mile delivery and logistics, among others. With recent progress in high-performance actuators, the hardware design of legged robot becomes relatively mature, while perceptive locomotion control algorithms still present significant challenges, preventing widespread deployment of legged robots in real-world applications. This talk will present recent developments in legged robots from both academic and industrial perspectives. Key research challenges in legged locomotion control will be identified. Both model-based and learning-based planning and control methods will be introduced, and their fundamental connections will be discussed. The results will be presented along with various experimental validations based on commercial platforms as well as legged robots custom designed and built by our lab. An Interview with R. Brockett Abstract: In this interview we explore the extraordinary career of Roger W. Brockett, An Wang Professor of Electrical Engineering and Computer Science, Emeritus at Harvard University. Here Roger reflects on his student days at Case Western University, his graduate study with M. Mesarovich and first job at MIT. The largest part of his career was spent on the faculty at Harvard University where he was an early initiator of geometric nonlinear control theory and went on to found the Harvard Robotics Laboratory. Roger also reflects on mentoring more than sixty PhD students - many of whom have become major figures in systems and control. Games, Decisions, and Control: Past, Present, and Future - Tamer Basar Abstract: Almost three-quarters of a century ago, when control was emerging as a major discipline, fueled by developments in dynamic, recursive decision making, and particularly dynamic programming, a separate discipline, differential games, was being created in response to the need to develop a framework and associated solution tools for strategic dynamic decision making in adversarial environments, as an outgrowth of game theory, still in its infancy at the time. The evolution of the two disciplines---control theory, and particularly optimal control, and the theory of differential/dynamic games (DGs)---initially followed somewhat different paths, with different foci, but soon a healthy interaction between the two developed. DGs, in both zero-sum and nonzero-sum settings, enabled the embedding of control in a broader context and framework, and enriched the set of conceptual tools available to it. Particularly, an essential ingredient of DGs, information structures (who knows what and when) entered and became a mainstay in control research. On the other hand, the field of DGs itself benefitted from the mathematical tools developed for control, such as the viscosity solutions. This interaction between the two reached a climax when the issue of robustness became prevalent in control, leading among others to a comprehensive treatment of H∞ control of both linear and nonlinear uncertain systems using the theory of zero-sum DGs with imperfect measurements. As a parallel development to the theory and applications of zero-sum DGs was that of nonzero-sum DGs, which required the introduction of additional tools and further enrichment of solution concepts, bringing along also many challenges in the characterization and computation of the associated solutions, arising because of complex strategic interactions among different entities (agents or players). More recently, as a major development, these challenges were partially overcome by going to infinite populations, creating the field of mean-field games, which again brought in strong connections to control theory, this time stochastic optimal control of systems driven also with exogenous stochastic processes. Again more recently, we see several learning tools developed in the context of adaptive control becoming relevant in multi-agent systems formulated as nonzero-sum DGs, for reinforcement learning within model-free settings. This presentation will dwell on the parallel developments in the two fields—control and dynamic/differential games—as they impacted each other over the past three-quarters of a century, talk about the present state of the art, and make projections into the future. Geometric Concepts in Nonlinear Control - Arthur Krener Abstract: We explore some geometric concepts that have enabled enormous progress in Nonlinear Control. They include the distributions that play a fundamental role in nonlinear controllability and observability. We also will talk about controlled and conditionally invariant distributions. These ideas are important in feedback linearization disturbance rejection and decoupling. The Center Manifold Theorem plays a crucial role in nonlinear regulation and brings us to topics like feedback stabilization and backstepping. If time permits we shall talk about normal forms of nonlinear control systems and control bifurcations. Geometric methods for control - A.S. Morse Abstract: More than a half century ago, two groups of researchers - one in Italy and the other in the United States - working independently, invented several key concepts which serve as the basis for what has become known as the theory of linear geometric control. The aim of this presentation is to explain what these concepts are, where they came from and why they are important. Robust Stabilizing Controllers with Robust Avoidance Properties (for linear systems with nontrivial drift) - Luca Zaccarian Workshop on Uncertain Dynamical Systems Consensus and Dissensus in Multi-population Multi-agent Systems - Tamer Basar Abstract: Perhaps the most challenging aspect of research on multi-agent dynamical systems, formulated as non-cooperative stochastic differential/dynamic games (SDGs) with asymmetric dynamic information structures, is the presence of strategic interactions among agents, with each one developing beliefs on others in the absence of shared information. This belief generation process involves what is known as second-guessing phenomenon, which generally entails infinite recursions, thus compounding the difficulty of obtaining (and arriving at) an equilibrium. This difficulty is somewhat alleviated when there is a high population of agents (players), in which case strategic interactions at the level of each agent become much less pronounced. With some structural specifications, this leads to what is known as mean field games (MFGs), which have been the subject of intense research activity during the last fifteen years or so. The talk will first provide a general overview of fundamentals of MFGs approach to decision making in multi-agent dynamical systems in both model-based and model-free settings, and discuss connections to finite-population games. Following this general introduction, the talk will focus, for concrete results, on the structured setting of discrete-time infinite-horizon linear-quadratic-Gaussian dynamic games, where the players are partitioned into finitely-many populations with an underlying graph topology---a framework motivated by paradigms where consensus and dissensus co-exist. In this formulation, each population has a high number of indistinguishable agents, but there is no indistinguishability across different populations. It is possible to characterize the Nash equilibrium (NE) of the underlying game when the number of agents in each population goes to infinity, the so-called mean-field equilibrium (MFE), with only local state information for each agent, which can then be shown to lead to an approximate NE when the population sizes are finite, with a precise quantification of the approximation as a function of population sizes. For the model-free versions of such games, a reinforcement learning algorithm will be introduced based on zero-order stochastic optimization, for computation of the MFE, along with guaranteed convergence. The talk will also address derivation of a finite-sample bound, quantifying the estimation error as a function of the number of samples, and will conclude with a discussion of some extensions of the general setting and future research directions. Capturing Carbon Economically, Today - Brett Henkel Mechatronic systems: past, present, future - Rolf Isermann Abstract: Many technical processes and products in the area of mechanical and electrical engineering show increasing integration of mechanics with digital electronics and information processing. This integration is between the components (hardware) and the information-driven functions (software), resulting in integrated systems called mechatronic systems. Their development involves finding an optimal balance between the basic mechanical structure, sensor and actuator implementation, and automatic information processing and overall control. Frequently formerly mechanical functions are replaced by electronically controlled functions, resulting in simpler mechanical structures and increased functionality. The development of mechatronic systems opens the door to many innovative solutions and synergetic effects which are not possible with mechanics or electronics alone. This technical progress has a very strong influence on a multitude of products in the areas of mechanical, electrical, and electronic engineering and is increasingly changing the design, for example, of conventional electromechanical components, machines, vehicles, and precision mechanical devices, and their connection to cloud services. The contribution describes besides of general aspects a systematic design of mechatronic systems, control software development and illustrates mechatronic developments for automobiles, including brake and steering systems and an example of nonlinear control of a high dynamic sports car with parameter and state estimation. NCCR Symposium on Socially Responsible Automation Presidents' Forum - Thomas Parisini - Moderator The Future of CSS Conferences and Publications Abstract: The 2022 CSS President will moderate this forum bringing together five CSS Past Presidents to debate and provide their vision on the future of CSS Conferences and Publications. Will a “new normal” arise for our conferences after restrictions due to the COVID-19 pandemic? Should we take the opportunity to improve our conferences by learning from the “provisional virtual mode” we experienced? Also, the COVID-19 pandemic has demonstrated how vital and impactful Open Access is. At the same time, it has also shown the inevitable key role played by the content certification provided through the peer review that learned societies and academic communities have supported for several decades. Will a sustainable Open Access model be possible for our journals? The forum will share opinions and landscapes in front of us. The Future of CSS Conferences and Publications Abstract: The 2022 CSS President will moderate this forum bringing together five CSS Past Presidents to debate and provide their vision on the future of CSS Conferences and Publications. Will a “new normal” arise for our conferences after restrictions due to the COVID-19 pandemic? Should we take the opportunity to improve our conferences by learning from the “provisional virtual mode” we experienced? Also, the COVID-19 pandemic has demonstrated how vital and impactful Open Access is. At the same time, it has also shown the inevitable key role played by the content certification provided through the peer review that learned societies and academic communities have supported for several decades. Will a sustainable Open Access model be possible for our journals? The forum will share opinions and landscapes in front of us. R.E. Kalman's legacy and sampled-data systems - Yutaka Yamamoto Abstract: This presentation gives a brief overview of R. E. Kalman's legacy. Kalman laid the foundation of modern control and system theory by giving many fundamental contributions. They constitute state space formalism, Kalman filtering, controllability/observability, optimal control and regulator theory, realization theory, sampled-data control, and so on. These contributions completely revolutionized control theory. We review their impact, and also attempt to clarify their conceptual backgrounds. We will discuss Kalman's philosophy that led to a fundamental paradigm shift from classical to modern. In particular, the basic concept of the state space formalism underlies all these contributions, and established modern control theory as a whole. As representative cases, we discuss in detail Kalman filtering, realization theory, and sampled-data systems, and clarify that the state space theory is the crucial key to the whole developments. In depth examination of his treatment of sampled-data systems naturally leads us to the consideration of intersample behavior, and this further leads us to a more modern development of sampled-data systems since the late 1980s. We also review this development as well as its outcomes, and then see some more recent results. We will conclude the talk by giving some indications for the future. Revisiting Self-tuning Regulators - Lei Guo Abstract: The self-tuning regulators (STR) are a basic class of adaptive controllers designed by combing the least-squares estimate with the minimum (tracking) variance control, which had played an important role in the development of adaptive control. Establishing a rigorous theory on the global stability and asymptotic optimality of the closed-loop stochastic systems under the STR had been a longstanding problem in control theory. This talk will give a retrospect of the research towards establishing a theory of STR and will present a detailed look at this theory for a basic class of linear stochastic systems. Revisiting STR may offer useful inspirations for investigating more complex adaptive (or intelligent) systems where machine learning is combined with feedback control. The ISS Philosophy as a Unifying Framework for Stability-like Behavior - Eduardo Sontag The legacy of Jan Willems - Arjan van der Schaft Abstract: The list of scientific contributions of Jan Willems spans a whole range of topics, including (input-output) stability theory, dissipativity theory, linear geometric control theory, identification, the behavioral approach to systems and control, physical systems theory, open stochastic systems, .. ; see the website https://homes.esat.kuleuven.be/~sistawww/smc/jwillems/ for much more information. In this talk I will focus on two of these topics: dissipativity theory and the behavioral approach to systems and control. Discussion of the first topic will be partly based on: Arjan van der Schaft, Introduction to Jan C. Willems’,“Dissipative Dynamical Systems, Part I: General Theory”: 50 Years of Dissipativity Theory, IEEE Control Systems Magazine 42 (2), 46-50, 2022, Part II: Linear Systems With Quadratic Supply Rates, IEEE Control Systems Magazine 42 (3), 32-35, 2022. YP and Industry Panel I - Rush Patel - Moderator YP and Industry Panel II - Rush Patel - Moderator Technology Showcase 2: Intelligent Transportation Systems Control of Intelligent Transportation Systems - Huei Peng Abstract: Intelligent transportation systems (ITS) is the application of sensing, analysis, control and communications technologies to ground transportation in order to improve safety, mobility and efficiency. From the late 1980’s, the focus and impact of control engineers (such as those associated with IEEE CSS) have evolved from active safety to automated driving. In this talk, we will review example systems where control theories and practice have been applied with great success in the last 40 years. This will be followed by talking about the main challenges and opportunities in the development of automated and autonomous systems. Technology Showcase 3: Automotive The long voyage towards autonomous driving, with control systems as the co-pilot - Stefano Di Cairano Abstract: Autonomous driving inspired many of our dreams as kids, and in today’s news we seem closer and closer to realize them. From the DARPA challenges twenty years ago, the technology has been moving forward significantly, both in terms of prototype vehicle capabilities and deployed features in production vehicles. Yet, before the automated cars of our youth dreams become reality, there are more challenges to address. On this long journey, control systems research may provide more than technical tools for development. It may guide us with a framework, a mindset and a design philosophy rooted in rigorous analysis, architecture structuring, operational guarantees, and adaptation to changing environments. I will take this opportunity to share some of my encounters with automated driving technologies, from the inspiration of the DARPA challenges, to the development of advanced driving assistance systems, to research on future automated driving, and how control shaped the way leading to such developments, some of which now are in mass production and available to the public. From the Engine to the Battery, a Twenty-year Powertrain Electrification Journey - Maarten Steinbuch, Theo Hofman, Mauro Salazar Abstract: Since the first introduction of hybrid electric cars by Toyota two decades ago, we have been witnessing a major shift with respect to the propulsion of vehicles. The intelligence added by the increase in sensing, computing and actuating capabilities within engines has enabled the implementation of sophisticated control mechanisms that made them much more clean, powerful and reliable, whilst also enabling them to cope with more and more stringent emissions regulations. At the same time, the combination of combustion engines with electric drives enabled the design and deployment of hybrid powertrains, reaching significantly lower fuel consumption and emission levels also thanks to the implementation of intelligent supervisory energy management systems. As we are now converging to battery electric drives, control system technologies for battery management systems are gaining momentum. By the end of this decade, electric powertrains are expected to be the standard for automotive, ending a period of more that 100 years of combustion engine development. Overall, this talk will highlight the prominent role of digitization and intelligence in vehicle propulsion systems and their journey towards a future that is smart and clean! Technology Showcase 4: Control for Societal-Scale Challenges: Roadmap 2030 - Anuradha M. Annaswamy, Karl H. Johansson, George Pappas Abstract: The field of control systems is focused on principles and methods that ensure desired performance by automatic monitoring, processing, and executing suitable decisions. Their development has been both broad and deep. The scope of application has been continuously increasing over the past few decades, moving from feedback control of a single device and a single system to large-scale systems, system of systems, and infrastructure systems. All of these advances and transformations urge a shift in the conversation towards grand societal-scale challenges and the role of control systems in this broad canvas. Prompted by a CSS initiative, a roadmap document is being put together through the efforts over 60 leaders in the area with a forward-looking goal of investigating societal imperatives on a global scale that the controls community can shape and influence. In particular, the roadmap will identify new and emerging directions in control theory and technology that serve as glide paths towards this goal. Two preparatory workshops held in 2021 (virtually) and 2022 (in Stockholm), under the sponsorship of CSS and US National Science Foundation, have paved the way for the preparation of the roadmap. Target audiences include the entire controls community, young researchers, and funding agencies. On the CSS Day, we will present some highlights of the roadmap. This will include (1) Societal Drivers, (2) Technological Trends, (3) Methodological Challenges, (4) Technology Validation and Transition, and (5) Education and Training. The presentation will be followed by a question and answer session. IEEE Kharagpur CSS Chapter Presentations Sixty Years of Control Theory in Sixty Minutes and Sixty Slides - M. Vidyasagar Q-learning and Its Variants for the Single and Multi-agent Setting - Shalabh Bhatnagar Adaptive Gradient Optimizer in Machine Learning: A Control Theoretic Perspective - Nikhil Chopra Contraction Theory for Network Systems - Francesco Bullo IEEE CSS Kansai Chapter Congruences and Quotients: A Model-Independent Structural Approach in Systems and Control - Kai Cai Abstract: An effective approach to understanding large complex dynamic systems is to reduce their sizes, or decompose their monolithic structures into simpler components. In this talk, we introduce an algebraic framework for property-preserving structural reduction and decomposition based on congruences, i.e. binary relations that respect system dynamics. Given a dynamic system with a property of interest, we construct a congruence for the property and show that the corresponding quotient yields a reduced or decomposed structure which preserves that property. We illustrate the generality of this framework with two existing methods: bisimulation based abstraction in hybrid control and supervisor localization in discrete-event systems. Finally we introduce an application of this framework in cyber-physical control of multi-robot systems.
