Searcharxiv⌕ Search

arXiv subjects

Sandeep Banik

Publications and source records attributed to Sandeep Banik.

12 recordsLinked to original sources
arXiv2026

Principled Authority Switching for Shared Autonomy in Human-Robot Teams

Shared autonomy requires principled mechanisms for allocating and transferring control between a human and an autonomous agent. Existing approaches often rely on blending control inputs or heuristic switching rules, which lack theoretical guarantees and fail to account for the dynamics of authority transfer. This paper develops a cooperative game-theoretic framework for authority switching in shared autonomy. We formulate the control switching problem as an identical-interest dynamic game in which authority transitions are embedded into the system dynamics, yielding optimal switching policies rather than ad hoc rules. We establish the existence and characterization of team-optimal policies in pure strategies under stochastic human override, accounting for asymmetric authority where humans retain override capability. For linear-quadratic systems, we derive closed-form recursions for the optimal switching policies and value functions, enabling efficient computation independent of the continuous state. We validate the framework on scalar and multi-dimensional linear systems, demonstrating how optimal switching adapts to varying system dynamics, cost structures, and override probabilities. The results reveal fundamental trade-offs between human adaptability and autonomous efficiency, illustrating the practical benefits of grounding shared autonomy in cooperative game theory.

cs.HC↗
arXiv2026

$\texttt{Flip-Team}$: Cooperative Takeover Games with Stochastic Human Override

Shared autonomy requires principled mechanisms for allocating and transferring control between a human and an autonomous agent. Existing approaches often rely on blending control inputs or heuristic switching rules, which lack theoretical guarantees and fail to account for the dynamics of authority transfer. This paper develops a cooperative game-theoretic framework for authority switching in shared autonomy. We formulate the control switching problem as an identical-interest dynamic game in which authority transitions are embedded into the system dynamics, yielding optimal switching policies rather than ad hoc rules. We establish the existence and characterization of team-optimal policies in pure strategies under stochastic human override, accounting for asymmetric authority where humans retain override capability. For linear-quadratic systems, we derive closed-form recursions for the optimal switching policies and value functions, enabling efficient computation independent of the continuous state. We validate the framework on scalar and multi-dimensional linear systems, demonstrating how optimal switching adapts to varying system dynamics, cost structures, and override probabilities. The results reveal fundamental trade-offs between human adaptability and autonomous efficiency, illustrating the practical benefits of grounding shared autonomy in cooperative game theory.

cs.HC↗
arXiv2025

Distributionally Robust Imitation Learning: Layered Control Architecture for Certifiable Autonomy

Imitation learning (IL) enables autonomous behavior by learning from expert demonstrations. While more sample-efficient than comparative alternatives like reinforcement learning, IL is sensitive to compounding errors induced by distribution shifts. There are two significant sources of distribution shifts when using IL-based feedback laws on systems: distribution shifts caused by policy error and distribution shifts due to exogenous disturbances and endogenous model errors due to lack of learning. Our previously developed approaches, Taylor Series Imitation Learning (TaSIL) and $\mathcal{L}_1$ -Distributionally Robust Adaptive Control (\ellonedrac), address the challenge of distribution shifts in complementary ways. While TaSIL offers robustness against policy error-induced distribution shifts, \ellonedrac offers robustness against distribution shifts due to aleatoric and epistemic uncertainties. To enable certifiable IL for learned and/or uncertain dynamical systems, we formulate \textit{Distributionally Robust Imitation Policy (DRIP)} architecture, a Layered Control Architecture (LCA) that integrates TaSIL and~\ellonedrac. By judiciously designing individual layer-centric input and output requirements, we show how we can guarantee certificates for the entire control pipeline. Our solution paves the path for designing fully certifiable autonomy pipelines, by integrating learning-based components, such as perception, with certifiable model-based decision-making through the proposed LCA approach.

eess.SY↗
arXiv2025

A Robust Task-Level Control Architecture for Learned Dynamical Systems

Dynamical system (DS)-based learning from demonstration (LfD) is a powerful tool for generating motion plans in the operation ('task') space of robotic systems. However, realizing generated motion plans is often compromised by a "task-execution mismatch", where unmodeled dynamics, persistent disturbances, and system latency cause the robot's task-space state to diverge from the desired state. We propose a novel task-level robust control architecture, L1-augmented Dynamical Systems (L1-DS), that explicitly handles the task-execution mismatch in tracking a nominal motion plan generated by any DS-based LfD scheme. Our framework augments any DS-based LfD model with a nominal stabilizing controller and an L1 adaptive controller. Furthermore, we introduce a windowed Dynamic Time Warping (DTW)-based target selector, which enables the nominal stabilizing controller to handle temporal misalignment for improved phase-consistent tracking. We demonstrate the efficacy of our architecture on the LASA and IROS handwriting datasets.

cs.RO↗
arXiv2025

Flip Co-op: Cooperative Takeovers in Shared Autonomy

Shared autonomy requires principled mechanisms for allocating and transferring control between a human and an autonomous agent. Existing approaches often rely on blending control inputs between human and autonomous agent or switching rules, which lack theoretical guarantees. This paper develops a game-theoretic framework for modeling cooperative takeover in shared autonomy. We formulate the switching interaction as a dynamic game in which authority is embedded directly into the system dynamics, resulting in Nash equilibrium(NE)-based strategies rather than ad hoc switching rules. We establish the existence and characterization of NE in the space of pure takeover strategies under stochastic human intent. For the class of linear-quadratic systems, we derive closed-form recursions for takeover strategies and saddle-point value functions, providing analytical insight and efficient computation of cooperative takeover policies. We further introduce a bimatrix potential game reformulation to address scenarios where human and autonomy utilities are not perfectly aligned, yielding a unifying potential function that preserves tractability while capturing intent deviations. The framework is applied to a vehicle trajectory tracking problem, demonstrating how equilibrium takeover strategies adapt across straight and curved path segments. The results highlight the trade-off between human adaptability and autonomous efficiency and illustrate the practical benefits of grounding shared autonomy in cooperative game theory.

cs.HC↗
arXiv2024

Integrating Vision Systems and STPA for Robust Landing and Take-Off in VTOL Aircraft

Vertical take-off and landing (VTOL) unmanned aerial vehicles (UAVs) are versatile platforms widely used in applications such as surveillance, search and rescue, and urban air mobility. Despite their potential, the critical phases of take-off and landing in uncertain and dynamic environments pose significant safety challenges due to environmental uncertainties, sensor noise, and system-level interactions. This paper presents an integrated approach combining vision-based sensor fusion with System-Theoretic Process Analysis (STPA) to enhance the safety and robustness of VTOL UAV operations during take-off and landing. By incorporating fiducial markers, such as AprilTags, into the control architecture, and performing comprehensive hazard analysis, we identify unsafe control actions and propose mitigation strategies. Key contributions include developing the control structure with vision system capable of identifying a fiducial marker, multirotor controller and corresponding unsafe control actions and mitigation strategies. The proposed solution is expected to improve the reliability and safety of VTOL UAV operations, paving the way for resilient autonomous systems.

cs.RO↗
arXiv2024

FlipDyn in Graphs: Resource Takeover Games in Graphs

We present \texttt{FlipDyn-G}, a dynamic game model extending the \texttt{FlipDyn} framework to a graph-based setting, where each node represents a dynamical system. This model captures the interactions between a defender and an adversary who strategically take over nodes in a graph to minimize (resp. maximize) a finite horizon additive cost. At any time, the \texttt{FlipDyn} state is represented as the current node, and each player can transition the \texttt{FlipDyn} state to a depending based on the connectivity from the current node. Such transitions are driven by the node dynamics, state, and node-dependent costs. This model results in a hybrid dynamical system where the discrete state (\texttt{FlipDyn} state) governs the continuous state evolution and the corresponding state cost. Our objective is to compute the Nash equilibrium of this finite horizon zero-sum game on a graph. Our contributions are two-fold. First, we model and characterize the \texttt{FlipDyn-G} game for general dynamical systems, along with the corresponding Nash equilibrium (NE) takeover strategies. Second, for scalar linear discrete-time dynamical systems with quadratic costs, we derive the NE takeover strategies and saddle-point values independent of the continuous state of the system. Additionally, for a finite state birth-death Markov chain (represented as a graph) under scalar linear dynamical systems, we derive analytical expressions for the NE takeover strategies and saddle-point values. We illustrate our findings through numerical studies involving epidemic models and linear dynamical systems with adversarial interactions.

cs.GT↗
arXiv2023

FlipDyn with Control: Resource Takeover Games with Dynamics

We introduce FlipDyn with control, a finite-horizon zero-sum resource takeover game, where a defender and an adversary decide when to takeover and how to control a common resource. At each discrete-time step, the players can take over or retain control, incurring state and control-dependent costs. The system is modeled as a hybrid dynamical system, with a discrete \texttt{FlipDyn} state determining control authority. Our contributions are: (i) For arbitrary non-negative costs, we derive the saddle-point value of the \texttt{FlipDyn} game and the corresponding Nash equilibria (NE) takeover strategies. (ii) For linear dynamical systems with quadratic costs, we establish sufficient conditions under which the game admits an NE. (iii) For scalar linear dynamical systems with quadratic costs, we derive parameterized NE takeover strategies and saddle-point values independent of the continuous state. (iv) For higher-dimensional linear dynamical systems with quadratic costs, we derive approximate NE takeover strategies and control policies, and compute bounds on the saddle-point values. We validate our results through a numerical study on adversarial control of a linear system.

eess.SY↗
arXiv2022

FlipDyn: A game of resource takeovers in dynamical systems

We introduce a game in which two players with opposing objectives seek to repeatedly takeover a common resource. The resource is modeled as a discrete time dynamical system over which a player can gain control after spending a state-dependent amount of energy at each time step. We use a FlipIT-inspired deterministic model that decides which player is in control at every time step. A player's policy is the probability with which the player should spend energy to gain control at each time step. Our main results are three-fold. First, we present analytic expressions for the cost-to-go as a function of the hybrid state of the system, i.e., the physical state of the dynamical system and the binary \texttt{FlipDyn} state for any general system with arbitrary costs. These expressions are exact when the physical state is also discrete and has finite cardinality. Second, for a continuous physical state with linear dynamics and quadratic costs, we derive expressions for Nash equilibrium (NE). For scalar physical states, we show that the NE depends only on the parameters of the value function and costs, and is independent of the state. Third, we derive an approximate value function for higher dimensional linear systems with quadratic costs. Finally, we illustrate our results through a numerical study on the problem of controlling a linear system in a given environment in the presence of an adversary.

cs.GT↗
arXiv2022

Secure Route Planning Using Dynamic Games with Stopping States

We consider the classic motion planning problem defined over a roadmap in which a vehicle seeks to find an optimal path from a source to a destination in presence of an attacker who can launch attacks on the vehicle over any edge of the roadmap. The vehicle (defender) has the capability to switch on/off a countermeasure that can detect and permanently disable the attack if it occurs concurrently. We model the problem of traveling along en edge using the framework of a simultaneous zero-sum dynamic game (edge-game) with a stopping state played between an attacker and defender. We characterize the Nash equiliria of an edge-game and provide closed form expressions for two actions per player. We further provide an analytic and approximate expression on the value of an edge-game and characterize conditions under which it grows sub-linearly with the number of stages. We study the sensitivity of Nash equilibrium to the (i) cost of using the countermeasure, (ii) cost of motion and (iii) benefit of disabling the attack. The solution of an edge-game is used to formulate and solve for the secure planning problem known as a meta-game. We design an efficient heuristic by converting the problem to a shortest path problem using the edge cost as the solution of corresponding edge-games. We illustrate our findings through several insightful simulations.

cs.GT↗
arXiv2020

Automated Adversary Emulation for Cyber-Physical Systems via Reinforcement Learning

Adversary emulation is an offensive exercise that provides a comprehensive assessment of a system's resilience against cyber attacks. However, adversary emulation is typically a manual process, making it costly and hard to deploy in cyber-physical systems (CPS) with complex dynamics, vulnerabilities, and operational uncertainties. In this paper, we develop an automated, domain-aware approach to adversary emulation for CPS. We formulate a Markov Decision Process (MDP) model to determine an optimal attack sequence over a hybrid attack graph with cyber (discrete) and physical (continuous) components and related physical dynamics. We apply model-based and model-free reinforcement learning (RL) methods to solve the discrete-continuous MDP in a tractable fashion. As a baseline, we also develop a greedy attack algorithm and compare it with the RL procedures. We summarize our findings through a numerical study on sensor deception attacks in buildings to compare the performance and solution quality of the proposed algorithms.

cs.LG↗
arXiv2019

Minimizing Age in Gateway Based Update Systems

We consider a network of status updating sensors whose updates are collected and sent to a monitor by a gateway. The monitor desires as fresh as possible updates from the network of sensors. The gateway may either poll a sensor for its status update or it may transmit collected sensor updates to the monitor. We derive the average age at the monitor for such a setting. We observe that increasing the frequency of transmissions to the monitor has the upside of resetting sensor age at the monitor to smaller values. However, it increases the length of time that elapses before a sensor is polled again. This motivates our investigation of policies that fix the number of sensors s the gateway polls before transmitting to the monitor. For any s, we show that when sensor transmission times to the gateway are independent and identically distributed (iid), for independent but possibly non-identical transmission times to the monitor, it is optimal to poll a sensor with the maximum age at the gateway first. Also, under simplifying assumptions, the optimal value of s increases as the square root of the number of sensors. For non-identical sensor transmission times, we consider a policy that polls a sensor such that the resulting average change in age is minimized. We compare our policy proposals with other policies, over a wide selection of transmission time distributions.

cs.IT↗