Jyotirmoy V. Deshmukh

Jyotirmoy V. Deshmukh, Xiaoqing Jin, Rupak Majumdar et al.

Embedded controllers for cyber-physical systems are often parameterized by look-up maps representing discretizations of continuous functions on metric spaces. For example, a non-linear control action may be represented as a table of pre-computed values, and the output action of the controller for a given input is computed by using interpolation. For industrial-scale control systems, several man-hours of effort is spent in tuning the values within the look-up maps, and sub-optimal performance is often associated with inappropriate values in look-up maps. Suppose that during testing, the controller code is found to have sub-optimal performance. The parameter fault localization problem asks which parameter values in the code are potential causes of the sub-optimal behavior. We present a statistical parameter fault localization approach based on binary similarity coefficients and set spectra methods. Our approach extends previous work on software fault localization to a quantitative setting where the parameters encode continuous functions over a metric space and the program is reactive. We have implemented our approach in a simulation workflow for automotive control systems in Simulink. Given controller code with parameters (including look-up maps), our framework bootstraps the simulation workflow to return a ranked list of map entries which are deemed to have most impact on the performance. On a suite of industrial case studies with seeded errors, our tool was able to precisely identify the location of the errors.

Navid Hashemi, Bardh Hoxha, Tomoya Yamaguchi et al.

Signal Temporal Logic (STL) has become a popular tool for expressing formal requirements of Cyber-Physical Systems (CPS). The problem of verifying STL properties of neural network-controlled CPS remains a largely unexplored problem. In this paper, we present a model for the verification of Neural Network (NN) controllers for general STL specifications using a custom neural architecture where we map an STL formula into a feed-forward neural network with ReLU activation. In the case where both our plant model and the controller are ReLU-activated neural networks, we reduce the STL verification problem to reachability in ReLU neural networks. We also propose a new approach for neural network controllers with general activation functions; this approach is a sound and complete verification approach based on computing the Lipschitz constant of the closed-loop control system. We demonstrate the practical efficacy of our techniques on a number of examples of learning-enabled control systems.

Navid Hashemi, Xin Qin, Jyotirmoy V. Deshmukh et al.

In this paper, we consider the problem of synthesizing a controller in the presence of uncertainty such that the resulting closed-loop system satisfies certain hard constraints while optimizing certain (soft) performance objectives. We assume that the hard constraints encoding safety or mission-critical task objectives are expressed using Signal Temporal Logic (STL), while performance is quantified using standard cost functions on system trajectories. In order to prioritize the satisfaction of the hard STL constraints, we utilize the framework of control barrier functions (CBFs) and algorithmically obtain CBFs for STL objectives. We assume that the controllers are modeled using neural networks (NNs) and provide an optimization algorithm to learn the optimal parameters for the NN controller that optimize the performance at a user-specified robustness margin for the safety specifications. We use the formalism of risk measures to evaluate the risk incurred by the trade-off between robustness margin of the system and its performance. We demonstrate the efficacy of our approach on well-known difficult examples for nonlinear control such as a quad-rotor and a unicycle, where the mission objectives for each system include hard timing constraints and safety objectives.

Xin Qin, Jyotirmoy V. Deshmukh

Producing probabilistic guarantee for several steps of a predicted signal follow a temporal logic defined behavior has its rising importance in monitoring. In this paper, we derive a method to compute the joint probability distribution of prediction errors of multiple steps based on Autoregressive Integrated Moving Average(ARIMA) model. We cover scenarios in stationary process and intrinsically stationary process for univariate and multivariate.

Sheryl Paul, Leslie Cruz Juarez, Jyotirmoy V. Deshmukh et al.

Networked multi-agent dynamical systems have been used to model how individual opinions evolve over time due to the opinions of other agents in the network. Particularly, such a model has been used to study how a planning agent can be used to steer opinions in a desired direction through repeated, budgeted interventions. In this paper, we consider the problem where individuals' susceptibilities to external influences are unknown. We propose an online algorithm that alternates between estimating this susceptibility parameter, and using the current estimate to drive the opinion to a desired target. We provide conditions that guarantee stability and convergence to the desired target opinion when the planning agent faces budgetary or temporal constraints. Our analysis shows that the key advantage of estimating the susceptibility parameter is that it helps achieve near-optimal convergence to the target opinion given a finite amount of intervention rounds, and, for a given intervention budget, quantifies how close the opinion can get to the desired target.

Yiqi Zhao, Xinyi Yu, Matteo Sesia et al.

We propose conformal predictive programming (CPP), a framework to solve chance constrained optimization problems, i.e., optimization problems with constraints that are functions of random variables. CPP utilizes samples from these random variables along with the quantile lemma - central to conformal prediction - to transform the chance constrained optimization problem into a deterministic problem with a quantile reformulation. CPP inherits a priori guarantees on constraint satisfaction from existing sample average approximation approaches for a class of chance constrained optimization problems, and it provides a posteriori guarantees that are of conditional and marginal nature otherwise. The strength of CPP is that it can easily support different variants of conformal prediction which have been (or will be) proposed within the conformal prediction community. To illustrate this, we present robust CPP to deal with distribution shifts in the random variables and Mondrian CPP to deal with class conditional chance constraints. To enable tractable solutions to the quantile reformulation, we present a mixed integer programming method (CPP-MIP) encoding, a bilevel optimization strategy (CPP-Bilevel), and a sampling-and-discarding optimization strategy (CPP-Discarding). We also extend CPP to deal with joint chance constrained optimization (JCCO). In a series of case studies, we show the validity of the aforementioned approaches, empirically compare CPP-MIP, CPP-Bilevel, as well as CPP-Discarding, and illustrate the advantage of CPP as compared to scenario approach.

Navid Hashemi, Bardh Hoxha, Danil Prokhorov et al.

This paper introduces a model-based approach for training feedback controllers for an autonomous agent operating in a highly nonlinear (albeit deterministic) environment. We desire the trained policy to ensure that the agent satisfies specific task objectives and safety constraints, both expressed in Discrete-Time Signal Temporal Logic (DT-STL). One advantage for reformulation of a task via formal frameworks, like DT-STL, is that it permits quantitative satisfaction semantics. In other words, given a trajectory and a DT-STL formula, we can compute the {\em robustness}, which can be interpreted as an approximate signed distance between the trajectory and the set of trajectories satisfying the formula. We utilize feedback control, and we assume a feed forward neural network for learning the feedback controller. We show how this learning problem is similar to training recurrent neural networks (RNNs), where the number of recurrent units is proportional to the temporal horizon of the agent's task objectives. This poses a challenge: RNNs are susceptible to vanishing and exploding gradients, and naïve gradient descent-based strategies to solve long-horizon task objectives thus suffer from the same problems. To tackle this challenge, we introduce a novel gradient approximation algorithm based on the idea of dropout or gradient sampling. One of the main contributions is the notion of {\em controller network dropout}, where we approximate the NN controller in several time-steps in the task horizon by the control input obtained using the controller in a previous training step. We show that our control synthesis methodology, can be quite helpful for stochastic gradient descent to converge with less numerical issues, enabling scalable backpropagation over long time horizons and trajectories over high dimensional state spaces.

Merve Atasever, Zhuochen Liu, Qingpei Li et al.

Autonomous driving remains a highly active research domain that seeks to enable vehicles to perceive dynamic environments, predict the future trajectories of traffic agents such as vehicles, pedestrians, and cyclists and plan safe and efficient future motions. To advance the field, several competitive platforms and benchmarks have been established to provide standardized datasets and evaluation protocols. Among these, leaderboards by the CARLA organization and nuPlan and the Waymo Open Dataset have become leading benchmarks for assessing motion planning algorithms. Each offers a unique dataset and challenging planning problems spanning a wide range of driving scenarios and conditions. In this study, we present a comprehensive comparative analysis of the motion planning methods featured on these three leaderboards. To ensure a fair and unified evaluation, we adopt CARLA leaderboard v2.0 as our common evaluation platform and modify the selected models for compatibility. By highlighting the strengths and weaknesses of current approaches, we identify prevailing trends, common challenges, and suggest potential directions for advancing motion planning research.

Francesca Cairoli, Luca Bortolussi, Jyotirmoy V. Deshmukh et al.

We consider the problem of quantitative predictive monitoring (QPM) of stochastic systems, i.e., predicting at runtime the degree of satisfaction of a desired temporal logic property from the current state of the system. Since computational efficiency is key to enable timely intervention against predicted violations, several state-of-the-art QPM approaches rely on fast machine-learning surrogates to provide prediction intervals for the satisfaction values, using conformal inference to offer statistical guarantees. However, these QPM methods suffer when the monitored agent exhibits multi-modal dynamics, whereby certain modes may yield high satisfaction values while others critically violate the property. Existing QPM methods are mode-agnostic and so would yield overly conservative and uninformative intervals that lack meaningful mode-specific satisfaction information. To address this problem, we present GenQPM, a method that leverages deep generative models, specifically score-based diffusion models, to reliably approximate the probabilistic and multi-modal system dynamics without requiring explicit model access. GenQPM employs a mode classifier to partition the predicted trajectories by dynamical mode. For each mode, we then apply conformal inference to produce statistically valid, mode-specific prediction intervals. We demonstrate the effectiveness of GenQPM on a benchmark of agent navigation and autonomous driving tasks, resulting in prediction intervals that are significantly more informative (less conservative) than mode-agnostic baselines.

Sheryl Paul, Jyotirmoy V. Deshmukh

Reinforcement learning (RL) has been successfully applied to solve the problem of finding obstacle-free paths for autonomous agents operating in stochastic and uncertain environments. However, when the underlying stochastic dynamics of the environment experiences drastic distribution shifts, the optimal policy obtained in the trained environment may be sub-optimal or may entirely fail in helping find goal-reaching paths for the agent. Approaches like domain randomization and robust RL can provide robust policies, but typically assume minor (bounded) distribution shifts. For substantial distribution shifts, retraining (either with a warm-start policy or from scratch) is an alternative approach. In this paper, we develop a novel approach called {\em Evolutionary Robust Policy Optimization} (ERPO), an adaptive re-training algorithm inspired by evolutionary game theory (EGT). ERPO learns an optimal policy for the shifted environment iteratively using a temperature parameter that controls the trade off between exploration and adherence to the old optimal policy. The policy update itself is an instantiation of the replicator dynamics used in EGT. We show that under fairly common sparsity assumptions on rewards in such environments, ERPO converges to the optimal policy in the shifted environment. We empirically demonstrate that for path finding tasks in a number of environments, ERPO outperforms several popular RL and deep RL algorithms (PPO, A3C, DQN) in many scenarios and popular environments. This includes scenarios where the RL algorithms are allowed to train from scratch in the new environment, when they are retrained on the new environment, or when they are used in conjunction with domain randomization. ERPO shows faster policy adaptation, higher average rewards, and reduced computational costs in policy adaptation.

Anand Balakrishnan, Stefan Jakšić, Edgar A. Aguilar et al.

Reinforcement learning (RL) is a popular approach for robotic path planning in uncertain environments. However, the control policies trained for an RL agent crucially depend on user-defined, state-based reward functions. Poorly designed rewards can lead to policies that do get maximal rewards but fail to satisfy desired task objectives or are unsafe. There are several examples of the use of formal languages such as temporal logics and automata to specify high-level task specifications for robots (in lieu of Markovian rewards). Recent efforts have focused on inferring state-based rewards from formal specifications; here, the goal is to provide (probabilistic) guarantees that the policy learned using RL (with the inferred rewards) satisfies the high-level formal specification. A key drawback of several of these techniques is that the rewards that they infer are sparse: the agent receives positive rewards only upon completion of the task and no rewards otherwise. This naturally leads to poor convergence properties and high variance during RL. In this work, we propose using formal specifications in the form of symbolic automata: these serve as a generalization of both bounded-time temporal logic-based specifications as well as automata. Furthermore, our use of symbolic automata allows us to define non-sparse potential-based rewards which empirically shape the reward surface, leading to better convergence during RL. We also show that our potential-based rewarding strategy still allows us to obtain the policy that maximizes the satisfaction of the given specification.

Mingxi Cheng, Junyao Zhang, Shahin Nazarian et al.

Many intelligent transportation systems are multi-agent systems, i.e., both the traffic participants and the subsystems within the transportation infrastructure can be modeled as interacting agents. The use of AI-based methods to achieve coordination among the different agents systems can provide greater safety over transportation systems containing only human-operated vehicles, and also improve the system efficiency in terms of traffic throughput, sensing range, and enabling collaborative tasks. However, increased autonomy makes the transportation infrastructure vulnerable to compromised vehicular agents or infrastructure. This paper proposes a new framework by embedding the trust authority into transportation infrastructure to systematically quantify the trustworthiness of agents using an epistemic logic known as subjective logic. In this paper, we make the following novel contributions: (i) We propose a framework for using the quantified trustworthiness of agents to enable trust-aware coordination and control. (ii) We demonstrate how to synthesize trust-aware controllers using an approach based on reinforcement learning. (iii) We comprehensively analyze an autonomous intersection management (AIM) case study and develop a trust-aware version called AIM-Trust that leads to lower accident rates in scenarios consisting of a mixture of trusted and untrusted agents.

Gaurav Gupta, Chenzhong Yin, Jyotirmoy V. Deshmukh et al.

Reinforcement learning (RL) is a technique to learn the control policy for an agent that interacts with a stochastic environment. In any given state, the agent takes some action, and the environment determines the probability distribution over the next state as well as gives the agent some reward. Most RL algorithms typically assume that the environment satisfies Markov assumptions (i.e. the probability distribution over the next state depends only on the current state). In this paper, we propose a model-based RL technique for a system that has non-Markovian dynamics. Such environments are common in many real-world applications such as in human physiology, biological systems, material science, and population dynamics. Model-based RL (MBRL) techniques typically try to simultaneously learn a model of the environment from the data, as well as try to identify an optimal policy for the learned model. We propose a technique where the non-Markovianity of the system is modeled through a fractional dynamical system. We show that we can quantify the difference in the performance of an MBRL algorithm that uses bounded horizon model predictive control from the optimal policy. Finally, we demonstrate our proposed framework on a pharmacokinetic model of human blood glucose dynamics and show that our fractional models can capture distant correlations on real-world datasets.

Aniruddh G. Puranic, Jyotirmoy V. Deshmukh, Stefanos Nikolaidis

Learning-from-demonstrations is an emerging paradigm to obtain effective robot control policies for complex tasks via reinforcement learning without the need to explicitly design reward functions. However, it is susceptible to imperfections in demonstrations and also raises concerns of safety and interpretability in the learned control policies. To address these issues, we use Signal Temporal Logic to evaluate and rank the quality of demonstrations. Temporal logic-based specifications allow us to create non-Markovian rewards, and also define interesting causal dependencies between tasks such as sequential task specifications. We validate our approach through experiments on discrete-world and OpenAI Gym environments, and show that our approach outperforms the state-of-the-art Maximum Causal Entropy Inverse Reinforcement Learning.

Parv Kapoor, Anand Balakrishnan, Jyotirmoy V. Deshmukh

Techniques based on Reinforcement Learning (RL) are increasingly being used to design control policies for robotic systems. RL fundamentally relies on state-based reward functions to encode desired behavior of the robot and bad reward functions are prone to exploitation by the learning agent, leading to behavior that is undesirable in the best case and critically dangerous in the worst. On the other hand, designing good reward functions for complex tasks is a challenging problem. In this paper, we propose expressing desired high-level robot behavior using a formal specification language known as Signal Temporal Logic (STL) as an alternative to reward/cost functions. We use STL specifications in conjunction with model-based learning to design model predictive controllers that try to optimize the satisfaction of the STL specification over a finite time horizon. The proposed algorithm is empirically evaluated on simulations of robotic system such as a pick-and-place robotic arm, and adaptive cruise control for autonomous vehicles.

Sara Mohammadinejad, Brandon Paulsen, Chao Wang et al.

Recurrent neural networks (RNNs) such as Long Short Term Memory (LSTM) networks have become popular in a variety of applications such as image processing, data classification, speech recognition, and as controllers in autonomous systems. In practical settings, there is often a need to deploy such RNNs on resource-constrained platforms such as mobile phones or embedded devices. As the memory footprint and energy consumption of such components become a bottleneck, there is interest in compressing and optimizing such networks using a range of heuristic techniques. However, these techniques do not guarantee the safety of the optimized network, e.g., against adversarial inputs, or equivalence of the optimized and original networks. To address this problem, we propose DIFFRNN, the first differential verification method for RNNs to certify the equivalence of two structurally similar neural networks. Existing work on differential verification for ReLUbased feed-forward neural networks does not apply to RNNs where nonlinear activation functions such as Sigmoid and Tanh cannot be avoided. RNNs also pose unique challenges such as handling sequential inputs, complex feedback structures, and interactions between the gates and states. In DIFFRNN, we overcome these challenges by bounding nonlinear activation functions with linear constraints and then solving constrained optimization problems to compute tight bounding boxes on nonlinear surfaces in a high-dimensional space. The soundness of these bounding boxes is then proved using the dReal SMT solver. We demonstrate the practical efficacy of our technique on a variety of benchmarks and show that DIFFRNN outperforms state-of-the-art RNN verification tools such as POPQORN.

Sara Mohammadinejad, Jyotirmoy V. Deshmukh, Aniruddh G. Puranic

Many complex cyber-physical systems can be modeled as heterogeneous components interacting with each other in real-time. We assume that the correctness of each component can be specified as a requirement satisfied by the output signals produced by the component, and that such an output guarantee is expressed in a real-time temporal logic such as Signal Temporal Logic (STL). In this paper, we hypothesize that a large subset of input signals for which the corresponding output signals satisfy the output requirement can also be compactly described using an STL formula that we call the environment assumption. We propose an algorithm to mine such an environment assumption using a supervised learning technique. Essentially, our algorithm treats the environment assumption as a classifier that labels input signals as good if the corresponding output signal satisfies the output requirement, and as bad otherwise. Our learning method simultaneously learns the structure of the STL formula as well as the values of the numeric constants appearing in the formula. To achieve this, we combine a procedure to systematically enumerate candidate Parametric STL (PSTL) formulas, with a decision-tree based approach to learn parameter values. We demonstrate experimental results on real world data from several domains including transportation and health care.

Kolby Nottingham, Anand Balakrishnan, Jyotirmoy Deshmukh et al.

It is notoriously difficult to control the behavior of reinforcement learning agents. Agents often learn to exploit the environment or reward signal and need to be retrained multiple times. The multi-objective reinforcement learning (MORL) framework separates a reward function into several objectives. An ideal MORL agent learns to generalize to novel combinations of objectives allowing for better control of an agent's behavior without requiring retraining. Many MORL approaches use a weight vector to parameterize the importance of each objective. However, this approach suffers from lack of expressiveness and interpretability. We propose using propositional logic to specify the importance of multiple objectives. By using a logic where predicates correspond directly to objectives, specifications are inherently more interpretable. Additionally the set of specifications that can be expressed with formal languages is a superset of what can be expressed by weight vectors. In this paper, we define a formal language based on propositional logic with quantitative semantics. We encode logical specifications using a recurrent neural network and show that MORL agents parameterized by these encodings are able to generalize to novel specifications over objectives and achieve performance comparable to single objective baselines.

Sara Mohammadinejad, Jyotirmoy V. Deshmukh, Aniruddh G. Puranic et al.

Cyber-physical system applications such as autonomous vehicles, wearable devices, and avionic systems generate a large volume of time-series data. Designers often look for tools to help classify and categorize the data. Traditional machine learning techniques for time-series data offer several solutions to solve these problems; however, the artifacts trained by these algorithms often lack interpretability. On the other hand, temporal logics, such as Signal Temporal Logic (STL) have been successfully used in the formal methods community as specifications of time-series behaviors. In this work, we propose a new technique to automatically learn temporal logic formulae that are able to cluster and classify real-valued time-series data. Previous work on learning STL formulas from data either assumes a formula-template to be given by the user, or assumes some special fragment of STL that enables exploring the formula structure in a systematic fashion. In our technique, we relax these assumptions, and provide a way to systematically explore the space of all STL formulas. As the space of all STL formulas is very large, and contains many semantically equivalent formulas, we suggest a technique to heuristically prune the space of formulas considered. Finally, we illustrate our technique on various case studies from the automotive, transportation and healthcare domain.

Cumhur Erkan Tuncali, James Kapinski, Hisahiro Ito et al.

We present a simulation-based approach for generating barrier certificate functions for safety verification of cyber-physical systems (CPS) that contain neural network-based controllers. A linear programming solver is utilized to find a candidate generator function from a set of simulation traces obtained by randomly selecting initial states for the CPS model. A level set of the generator function is then selected to act as a barrier certificate for the system, meaning it demonstrates that no unsafe system states are reachable from a given set of initial states. The barrier certificate properties are verified with an SMT solver. This approach is demonstrated on a case study in which a Dubins car model of an autonomous vehicle is controlled by a neural network to follow a given path.

Marcell Vazquez-Chanlatte, Shromona Ghosh, Jyotirmoy V. Deshmukh et al.

Cyber-physical systems of today are generating large volumes of time-series data. As manual inspection of such data is not tractable, the need for learning methods to help discover logical structure in the data has increased. We propose a logic-based framework that allows domain-specific knowledge to be embedded into formulas in a parametric logical specification over time-series data. The key idea is to then map a time series to a surface in the parameter space of the formula. Given this mapping, we identify the Hausdorff distance between boundaries as a natural distance metric between two time-series data under the lens of the parametric specification. This enables embedding non-trivial domain-specific knowledge into the distance metric and then using off-the-shelf machine learning tools to label the data. After labeling the data, we demonstrate how to extract a logical specification for each label. Finally, we showcase our technique on real world traffic data to learn classifiers/monitors for slow-downs and traffic jams.

Marcell Vazquez-Chanlatte, Jyotirmoy V. Deshmukh, Xiaoqing Jin et al.

To effectively analyze and design cyberphysical systems (CPS), designers today have to combat the data deluge problem, i.e., the burden of processing intractably large amounts of data produced by complex models and experiments. In this work, we utilize monotonic Parametric Signal Temporal Logic (PSTL) to design features for unsupervised classification of time series data. This enables using off-the-shelf machine learning tools to automatically cluster similar traces with respect to a given PSTL formula. We demonstrate how this technique produces interpretable formulas that are amenable to analysis and understanding using a few representative examples. We illustrate this with case studies related to automotive engine testing, highway traffic analysis, and auto-grading massively open online courses.

Jyotirmoy V. Deshmukh, Alexandre Donzé, Shromona Ghosh et al.

Signal Temporal Logic (STL) is a formalism used to rigorously specify requirements of cyberphysical systems (CPS), i.e., systems mixing digital or discrete components in interaction with a continuous environment or analog com- ponents. STL is naturally equipped with a quantitative semantics which can be used for various purposes: from assessing the robustness of a specification to guiding searches over the input and parameter space with the goal of falsifying the given property over system behaviors. Algorithms have been proposed and implemented for offline computation of such quantitative semantics, but only few methods exist for an online setting, where one would want to monitor the satisfaction of a formula during simulation. In this paper, we formalize a semantics for robust online monitoring of partial traces, i.e., traces for which there might not be enough data to decide the Boolean satisfaction (and to compute its quantitative counterpart). We propose an efficient algorithm to compute it and demonstrate its usage on two large scale real-world case studies coming from the automotive domain and from CPS education in a Massively Open Online Course (MOOC) setting. We show that savings in computationally expensive simulations far outweigh any overheads incurred by an online approach.

Jyotirmoy V. Deshmukh, Rupak Majumdar, Vinayak S. Prabhu

The conformance testing problem for dynamical systems asks, given two dynamical models (e.g., as Simulink diagrams), whether their behaviors are "close" to each other. In the semi-formal approach to conformance testing, the two systems are simulated on a large set of tests, and a metric, defined on pairs of real-valued, real-timed trajectories, is used to determine a lower bound on the distance. We show how the Skorkhod metric on continuous dynamical systems can be used as the foundation for conformance testing of complex dynamical models. The Skorokhod metric allows for both state value mismatches and timing distortions, and is thus well suited for checking conformance between idealized models of dynamical systems and their implementations. We demonstrate the robustness of the system conformance quantification by proving a \emph{transference theorem}: trajectories close under the Skorokhod metric satisfy "close" logical properties. Specifically, we show the result for the timed linear time logic \TLTL augmented with a rich class of temporal and spatial constraint predicates. We provide a window-based streaming algorithm to compute the Skorokhod metric, and use it as a basis for a conformance testing tool for Simulink. We experimentally demonstrate the effectiveness of our tool in finding discrepant behaviors on a set of control system benchmarks, including an industrial challenge problem.