Shankar Sastry

University of California, Berkeley | UCB · Department of Electrical Engineering and Computer Sciences

Tolling, or congestion pricing, has emerged as an effective tool for preventing gridlock in traffic systems. However, tolls are currently mostly designed on route-based traffic assignment models (TAM), which may be unrealistic and computationally expensive. Existing approaches also impractically assume that the central tolling authority can access...

How can the system operator learn an incentive mechanism that achieves social optimality based on limited information about the agents' behavior, who are dynamically updating their strategies? To answer this question, we propose an \emph{adaptive} incentive mechanism. This mechanism updates the incentives of agents based on the feedback of each age...

This paper proposes a new framework of Markov α-potential games to study Markov games. In this new framework, Markov games are shown to be Markov α-potential games, and the existence of an associated α-potential function is established. Any optimizer of an α-potential function is shown to be an α-stationary NE. Two important classes of practically...

Tolling in traffic networks offers a popular measure to minimize overall congestion. Existing toll designs primarily focus on congestion in route-based traffic assignment models (TAMs), in which travelers make a single route selection from their source to destination. However, these models do not reflect real-world traveler decisions because they p...

This work proposes an algorithm for explicitly constructing a pair of neural networks that linearize and reconstruct an embedded submanifold, from finite samples of this manifold. Our such-generated neural networks, called flattening networks (FlatNet), are theoretically interpretable, computationally feasible at scale, and generalize well to test...

Arc-based traffic assignment models (TAMs) are a popular framework for modeling traffic network congestion generated by self-interested travelers who sequentially select arcs based on their perceived latency on the network. However, existing arc-based TAMs either assign travelers to cyclic paths, or do not extend to networks with bi-directional arc...

The rapid technological growth observed in the sUAS sector over the past decade has been unprecedented and has left gaps in policies and regulations to adequately provide for a safe and trusted environment in which to operate these devices. The Center for Security in Politics at UC Berkeley, via a two-day workshop, analyzed these gaps by addressing...

Causal phenomena associated with rare events frequently occur across a wide range of engineering and mathematical problems, such as risk-sensitive safety analysis, accident analysis and prevention, and extreme value theory. However, current methods for causal discovery are often unable to uncover causal links between random variables that manifest...

We study the problem of online learning in two-sided non-stationary matching markets, where the objective is to converge to a stable match. In particular, we consider the setting where one side of the market, the arms, has fixed known set of preferences over the other side, the players. While this problem has been studied when the players have fixe...

Recent advances in the reinforcement learning (RL) literature have enabled roboticists to automatically train complex policies in simulated environments. However, due to the poor sample complexity of these methods, solving reinforcement learning problems using real-world data remains a challenging problem. This paper introduces a novel cost-shaping...

Simultaneous Localization and Mapping (SLAM) algorithms perform visual-inertial estimation via filtering or batch optimization methods. Empirical evidence suggests that filtering algorithms are computationally faster, while optimization methods are more accurate. This work presents an optimization-based framework that unifies these approaches, and...

We study the problem of online learning in competitive settings in the context of two-sided matching markets. In particular, one side of the market, the agents, must learn about their preferences over the other side, the firms, through repeated interaction while competing with other agents for successful matches. We propose a class of decentralized...

We study the problem of online learning in competitive settings in the context of two-sided matching markets. In particular, one side of the market, the agents, must learn about their preferences over the other side, the firms, through repeated interaction while competing with other agents for successful matches. We propose a class of decentralized...

We propose a multi-agent reinforcement learning dynamics, and analyze its convergence properties in infinite-horizon discounted Markov potential games. We focus on the independent and decentralized setting, where players can only observe the realized state and their own reward in every stage. Players do not have knowledge of the game model, and can...

How can a social planner adaptively incentivize selfish agents who are learning in a strategic environment to induce a socially optimal outcome in the long run? We propose a two-timescale learning dynamics to answer this question in both atomic and non-atomic games. In our learning dynamics, players adopt a class of learning rules to update their s...

Optimal control is an essential tool for stabilizing complex nonlinear system. However, despite the extensive impacts of methods such as receding horizon control, dynamic programming and reinforcement learning, the design of cost functions for a particular system often remains a heuristic-driven process of trial and error. In this paper we seek to...

Decentralized planning for multi-agent systems, such as fleets of robots in a search-and-rescue operation, is often constrained by limitations on how agents can communicate with each other. One such limitation is the case when agents can communicate with each other only when they are in line-of-sight (LOS). Developing decentralized planning methods...

Simultaneous Localization and Mapping (SLAM) algorithms perform visual-inertial estimation via filtering or batch optimization methods. Empirical evidence suggests that filtering algorithms are computationally faster, while optimization methods are more accurate. This work presents an optimization-based framework that unifies these approaches, and...

How to design tolls that induce socially optimal traffic loads with dynamically arriving travelers who make selfish routing decisions? We propose a two-timescale discrete-time stochastic dynamics that adaptively adjusts the toll prices on a parallel link network while accounting for the updates of traffic loads induced by the incoming and outgoing...

This paper investigates optimal control problems formulated over a class of hybrid dynamical systems which display event-triggered discrete jumps. Due to the discontinuous nature of the underlying dynamics, previous approaches to solving optimal control problems over this class of systems generally rely on fixing the number and sequence of discrete...

As predictive models are deployed into the real world, they must increasingly contend with strategic behavior. A growing body of work on strategic classification treats this problem as a Stackelberg game: the decision-maker "leads" in the game by deploying a model, and the strategic agents "follow" by playing their best response to the deployed mod...

As predictive models are deployed into the real world, they must increasingly contend with strategic behavior. A growing body of work on strategic classification treats this problem as a Stackelberg game: the decision-maker "leads" in the game by deploying a model, and the strategic agents "follow" by playing their best response to the deployed mod...

Min-max optimization is emerging as a key framework for analyzing problems of robustness to strategically and adversarially generated data. We propose a random reshuffling-based gradient free Optimistic Gradient Descent-Ascent algorithm for solving convex-concave min-max problems with finite sum structure. We prove that the algorithm enjoys the sam...

Min-max optimization is emerging as a key framework for analyzing problems of robustness to strategically and adversarially generated data. We propose a random reshuffling-based gradient free Optimistic Gradient Descent-Ascent algorithm for solving convex-concave min-max problems with finite sum structure. We prove that the algorithm enjoys the sam...

The widespread adoption of nonlinear Receding Horizon Control (RHC) strategies by industry has led to more than 30 years of intense research efforts to provide stability guarantees for these methods. However, current theoretical guarantees require that each (generally nonconvex) planning problem can be solved to (approximate) global optimality, whi...

When an expert operates a perilous dynamic system, ideal constraint information is tacitly contained in their demonstrated trajectories and controls. The likelihood of these demonstrations can be computed, given the system dynamics and task objective, and the maximum likelihood constraints can be identified. Prior constraint inference work has focu...

This paper investigates optimal control problems formulated over a class of hybrid dynamical systems which display event-triggered discrete jumps. Due to the discontinuous nature of the underlying dynamics, previous approaches to solving optimal control problems over this class of systems generally rely on fixing the number and sequence of discrete...

This paper introduces a framework for learning a safe, stabilizing controller for a system with unknown dynamics using model-free policy optimization algorithms. Using a nominal dynamics model, the user specifies a candidate Control Lyapunov Function (CLF) around the desired operating point, and specifies the desired safe-set using a Control Barrie...

We analyze stochastic differential equations and their discretizations to derive novel high probability tracking bounds for exponentially stable time varying systems which are corrupted by process noise. The bounds have an explicit dependence on the rate of convergence for the unperturbed system and the dimension of the state space. The magnitude o...

In this work we present a multi-armed bandit framework for online expert selection in Markov decision processes and demonstrate its use in high-dimensional settings. Our method takes a set of candidate expert policies and switches between them to rapidly identify the best performing expert using a variant of the classical upper confidence bound alg...

The following topics are dealt with: control system synthesis; nonlinear control systems; linear systems; stability; optimisation; feedback; closed loop systems; Lyapunov methods; multi-agent systems; optimal control.

In this work we present a multi-armed bandit framework for online expert selection in Markov decision processes and demonstrate its use in high-dimensional settings. Our method takes a set of candidate expert policies and switches between them to rapidly identify the best performing expert using a variant of the classical upper confidence bound alg...

Collaboration requires coordination, and we coordinate by anticipating our teammates’ future actions and adapting to their plan. In some cases, our teammates’ actions early on can give us a clear idea of what the remainder of their plan is, i.e. what action sequence we should expect. In others, they might leave us less confident, or even lead us to...

We present a novel approach to control design for nonlinear systems which leverages model-free policy optimization techniques to learn a linearizing controller for a physical plant with unknown dynamics. Feedback linearization is a technique from nonlinear control which renders the input-output dynamics of a nonlinear plant linear under application...

This paper introduces a framework for learning a minimum-norm stabilizing controller for a system with unknown dynamics using model-free policy optimization methods. The approach begins by first designing a Control Lyapunov Function (CLF) for a (possibly inaccurate) dynamics model for the system, along with a function which specifies a minimum acce...

The main drawbacks of input-output linearizing controllers are the need for precise dynamics models and not being able to account for input constraints. Model uncertainty is common in almost every robotic application and input saturation is present in every real world system. In this paper, we address both challenges for the specific case of bipeda...

This paper proposes a framework for adaptively learning a feedback linearization-based tracking controller for an unknown system using discrete-time model-free policy-gradient parameter update rules. The primary advantage of the scheme over standard model-reference adaptive control techniques is that it does not require the learned inverse model to...

We present a novel first order controller for systems evolving on matrix Lie groups, a major use case of which is Cartesian velocity control on robot manipulators. This controller achieves global exponential trajectory tracking on a number of commonly used Lie groups including the Special Orthogonal Group SO(n), the Special Euclidean Group SE(n), a...

We introduce a general framework for competitive gradient-based learning that encompasses a wide breadth of multiagent learning algorithms, and analyze the limiting behavior of competitive gradient-based learning algorithms using dynamical systems theory. For both general-sum and potential games, we characterize a nonnegligible subset of the local...

Robots need models of human behavior for both inferring human goals and preferences, and predicting what people will do. A common model is the Boltzmann noisily-rational decision model, which assumes people approximately optimize a reward function and choose trajectories in proportion to their exponentiated reward. While this model has been success...

When an online learning algorithm is used to estimate the unknown parameters of a model, the signals interacting with the parameter estimates should not decay too quickly for the optimal values to be discovered correctly. This requirement is referred to as persistency of excitation, and it arises in various contexts, such as optimization with stoch...

As intelligent systems gain autonomy and capability, it becomes vital to ensure that their objectives match those of their human users; this is known as the value-alignment problem. In robotics, value alignment is key to the design of collaborative robots that can integrate into human workflows, successfully inferring and adapting to their users’ o...

We present a novel approach to control design for nonlinear systems, which leverages reinforcement learning techniques to learn a linearizing controller for a physical plant with unknown dynamics. Feedback linearization is a technique from nonlinear control which renders the input-output dynamics of a nonlinear plant \emph{linear} under application...

While most approaches to the problem of Inverse Reinforcement Learning (IRL) focus on estimating a reward function that best explains an expert agent's policy or demonstrated behavior on a control task, it is often the case that such behavior is more succinctly described by a simple reward combined with a set of hard constraints. In this setting, t...

We show by counterexample that policy-gradient algorithms have no guarantees of even local convergence to Nash equilibria in continuous action and state space multi-agent settings. To do so, we analyze gradient-play in $N$-player general-sum linear quadratic games. In such games the state and action spaces are continuous and the unique global Nash...

We show by counterexample that policy-gradient algorithms have no guarantees of even local convergence to Nash equilibria in continuous action and state space multi-agent settings. To do so, we analyze gradient-play in N-player general-sum linear quadratic games, a classic game setting which is recently emerging as a benchmark in the field of multi...

In recent years, data have played an increasingly important role in the economy as a good in its own right. In many settings, data aggregators cannot directly verify the quality of the data they purchase, nor the effort exerted by data sources when creating the data. Recent work has explored mechanisms to ensure that the data sources share high-qua...

Emerging industrial platforms such as the Internet of Things (IoT), Industrial Internet (II) in the US and Industrie 4.0 in Europe have tremendously accelerated the development of new generations of Cyber-Physical Systems (CPS) that integrate humans and human organizations (H-CPS) with physical and computation processes and extend to societal-scale...

In recent years, data has played an increasingly important role in the economy as a good in its own right. In many settings, data aggregators cannot directly verify the quality of the data they purchase, nor the effort exerted by data sources when creating the data. Recent work has explored mechanisms to ensure that the data sources share high qual...

This paper investigates optimal control problems formulated over a class of piecewise-smooth vector fields. Instead of optimizing over the discontinuous system directly, we instead formulate optimal control problems over a family of regularizations which are obtained by "smoothing out" the discontinuity in the original system. It is shown that the...

This paper investigates optimal control problems formulated over a class of piecewise-smooth vector fields. Instead of optimizing over the discontinuous system directly, we instead formulate optimal control problems over a family of regularizations which are obtained by "smoothing out" the discontinuity in the original system. It is shown that the...

State-of-the-art neural networks are vulnerable to adversarial examples; they can easily misclassify inputs that are imperceptibly different than their training and test data. In this work, we establish that the use of cross-entropy loss function and the low-rank features of the training data have responsibility for the existence of these inputs. B...

We propose a two-timescale algorithm for finding local Nash equilibria in two-player zero-sum games. We first show that previous gradient-based algorithms cannot guarantee convergence to local Nash equilibria due to the existence of non-Nash stationary points. By taking advantage of the differential structure of the game, we construct an algorithm...

We propose local symplectic surgery, a two-timescale procedure for finding local Nash equilibria in two-player zero-sum games. We first show that previous gradient-based algorithms cannot guarantee convergence to local Nash equilibria due to the existence of non-Nash stationary points. By taking advantage of the differential structure of the game,...

As human-robot systems make their ways into our every day life, safety has become a core concern of the learning algorithms used by such systems. Examples include semi-autonomous vehicles such as automobiles and aircrafts. The robustness of controllers in such systems relies on the accuracy of models of human behavior. In this paper, we propose a s...

The actions of an autonomous vehicle on the road affect and are affected by those of other drivers, whether overtaking, negotiating a merge, or avoiding an accident. This mutual dependence, best captured by dynamic game theory, creates a strong coupling between the vehicle's planning and its predictions of other drivers' behavior, and constitutes a...

Traditionally, autonomous cars treat human-driven vehicles like moving obstacles. They predict their future trajectories and plan to stay out of their way. While physically safe, this results in defensive and opaque behaviors. In reality, an autonomous car’s actions will actually affect what other cars will do in response, creating an opportunity f...

The Internet of Things (IoT) promises many advantages in the control and monitoring of physical systems from both efficacy and efficiency perspectives. However, in the wrong hands, the data might pose a privacy threat. In this article, we consider the tradeoff between the operational value of data collected in the IoT and the privacy of consumers....

Training a neural network with the gradient descent algorithm gives rise to a discrete-time nonlinear dynamical system. Consequently, behaviors that are typically observed in these systems emerge during training, such as convergence to an orbit but not to a fixed point or dependence of convergence on the initialization. Step size of the algorithm p...

When a human supervisor collaborates with a team of robots, their attention is divided and cognitive resources are at a premium. We aim to optimize the distribution of these resources and the flow of attention. To this end, we propose the model of an idealized supervisor to describe human behavior. Such a supervisor employs a potentially inaccurate...

Hybrid dynamical systems have proven to be a powerful modeling abstraction, yet fundamental questions regarding their dynamical properties remain. In this paper, we develop a novel solution concept for a class of hybrid systems, which is a generalization of Filippov's solution concept. In the mathematical theory, these \emph{hybrid Filippov solutio...

While training error of most deep neural networks degrades as the depth of the network increases, residual networks appear to be an exception. We show that the main reason for this is the Lyapunov stability of the gradient descent algorithm: for an arbitrarily chosen step size, the equilibria of the gradient descent are most likely to remain stable...

Collaboration requires coordination, and we coordinate by anticipating our teammates' future actions and adapting to their plan. In some cases, our teammates' actions early on can give us a clear idea of what the remainder of their plan is, i.e. what action sequence we should expect. In others, they might leave us less confident, or even lead us to...

The study of human-robot interaction is fundamental to the design and use of robotics in real-world applications. Robots will need to predict and adapt to the actions of human collaborators in order to achieve good performance and improve safety and end-user adoption. This paper evaluates a human-robot collaboration scheme that combines the task al...

We address the problem of inverse reinforcement learning in Markov decision processes where the agent is risksensitive. In particular, we model risk-sensitivity in a reinforcement learning framework by making use of models of human decision-making having their origins in behavioral psychology and economics. We propose a gradient-based inverse reinf...

Despite growing attention in autonomy, there are still many open problems, including how autonomous vehicles will interact and communicate with other agents, such as human drivers and pedestrians. Unlike most approaches that focus on pedestrian detection and planning for collision avoidance, this paper considers modeling the interaction between hum...

Hybrid dynamical systems have proven to be a powerful modeling abstraction, yet fundamental questions regarding the dynamical properties of these systems remain. In this paper, we develop a novel class of relaxations which we use to recover a number of classic systems theoretic properties for hybrid systems, such as existence and uniqueness of traj...