arXiv subjects

Shankar Sastry

Publications and source records attributed to Shankar Sastry.

At least 19 recordsLinked to original sources

arXiv2026

Congestion Games with Heterogeneous Valuations: An Optimal Transport Approach

In emerging urban mobility and logistics applications, such as advanced air mobility, electric vehicle charging, and shared service systems, agents with heterogeneous valuations choose among multiple destinations while sharing congested network resources. However, existing congestion game and resource allocation models do not simultaneously capture heterogeneous destination preferences and aggregate congestion externalities. We introduce a new nonatomic congestion game framework in which agents are endowed with heterogeneous destination valuations modeled by a measure space. We characterize Nash equilibria and social optima in this setting and show that both admit finite-dimensional representations in terms of threshold vectors and dual potentials. These structures induce a partition of the valuation space that determines agents' destination choices. Our analysis leverages Kantorovich duality from optimal transport theory and provides a new geometric perspective on congestion games with heterogeneous valuations.

cs.GT↗

arXiv2026

Hierarchical Generative Agents for Simulating Sequential Human Behavior

Complex cognitive, emotional, and social processes shape human evacuations during natural disasters. Accurate modeling and understanding of human behavior in disasters or emergencies can greatly impact the evacuation process by informing more effective planning and resource allocation. However, collecting human data in these situations is very difficult, and existing computational evacuation models assume rational, homogeneous behavior, leading to unrealistic, overly optimistic predictions. To address this gap, we present a simulation framework of sequential human decision-making during an evacuation scenario, introducing cognitively grounded, persona-driven agents. Our framework models evacuation behavior in a grid-based urban environment that evolves over time, capturing fire and other hazards. Human agents are modeled as personas that make sequential decisions in response to environmental stimuli with cognition structured in three levels: high-level evacuation goals, mid-level route reasoning, and low-level navigation. Decision-making is driven by large language models (LLMs) coupled with a cognitive module and calibrated with empirical human evacuation data. We propose a dynamic, stimulus-driven disaster simulation framework that models human evacuation decision-making using persona-conditioned LLM agents and a cognitive hierarchy.

cs.MA↗

arXiv2026

TT4D: A Pipeline and Dataset for Table Tennis 4D Reconstruction From Monocular Videos

We present TT4D, a large-scale, high-fidelity table tennis dataset. It provides $140+$ hours of reconstructed singles and doubles gameplay from monocular broadcast videos, featuring multimodal annotations like high-quality camera calibrations, precise 3D ball positions, ball spin, time segmentation, and 3D human meshes over time. This rich data provides a new foundation for virtual replay, in-depth player analysis, and robot learning. The dataset's combination of scale and precision is achieved through a novel reconstruction pipeline. Prior methods first partition a game sequence into individual shot segments based on the 2D ball track, and only then attempt reconstruction. However, 2D-based time segmentation collapses under occlusion and varied camera viewpoints, preventing reliable reconstruction. We invert this paradigm by first lifting the entire unsegmented 2D ball track to 3D through a learned lifting network. This 3D trajectory then allows us to reliably perform time segmentation. The learned lifting network also infers the ball's spin, handles unreliable ball detections, and successfully reconstructs the ball trajectory in cases of high occlusion. This lift-first design is necessary, as our pipeline is the only method capable of reconstructing table tennis gameplay from general-view broadcast monocular videos. We demonstrate the dataset's fidelity through two downstream tasks: estimating the racket's pose \& velocity at impact, and training a generative model of competitive rallies.

cs.CV↗

arXiv2026

Resilient Decentralized Ergodic Coverage for Scalable Multi-Robot Systems in Unknown Time-Varying Environments

Maintaining situational awareness in high-stakes multi-robot applications requires balancing exploration of unobserved regions with sustained monitoring of changing Regions of Interest (ROIs), often under unknown and time-varying distributions, partial observability, and limited communication. We propose a decentralized multi-agent coverage framework that serves as a high-level planning strategy, in which each agent computes an adaptive ergodic policy, implemented via a Markov-chain, that tracks an updated belief over the underlying importance map. Beliefs are maintained online via Gaussian Process (GP) regression from local noisy observations exchanged with neighbors. The resulting policy drives agents to spend time in ROIs in proportion to their estimated importance, while preserving sufficient exploration to detect and adapt to time-varying environmental changes. Unlike existing approaches that assume known importance maps, centralized coordination, or a static environment, our framework addresses the combined challenges of unknown, time-varying distributions under a decentralized, partially observable setting. We further show that our framework is robust to communication and memory degradation, robot loss, and can scale up to hundreds of robots.

cs.MA↗

arXiv2026

CaP-X: A Framework for Benchmarking and Improving Coding Agents for Robot Manipulation

"Code-as-Policy" considers how executable code can complement data-intensive Vision-Language-Action (VLA) methods, yet their effectiveness as autonomous controllers for embodied manipulation remains underexplored. We present CaP-X, an open-access framework for systematically studying Code-as-Policy agents in robot manipulation. At its core is CaP-Gym, an interactive environment in which agents control robots by synthesizing and executing programs that compose perception and control primitives. Building on this foundation, CaP-Bench evaluates frontier language and vision-language models across varying levels of abstraction, interaction, and perceptual grounding. Across 12 models, CaP-Bench reveals a consistent trend: performance improves with human-crafted abstractions but degrades as these priors are removed, exposing a dependence on designer scaffolding. At the same time, we observe that this gap can be mitigated through scaling agentic test-time computation--through multi-turn interaction, structured execution feedback, visual differencing, automatic skill synthesis, and ensembled reasoning--substantially improves robustness even when agents operate over low-level primitives. These findings allow us to derive CaP-Agent0, a training-free framework that recovers human-level reliability on several manipulation tasks in simulation and on real embodiments. We further introduce CaP-RL, showing reinforcement learning with verifiable rewards improves success rates and transfers from sim2real with minimal gap. Together, CaP-X provides a principled, open-access platform for advancing embodied coding agents.

cs.RO↗

arXiv2026

NePPO: Near-Potential Policy Optimization for General-Sum Multi-Agent Reinforcement Learning

Multi-agent reinforcement learning (MARL) is increasingly used to design learning-enabled agents that interact in shared environments. However, training MARL algorithms in general-sum games remains challenging: learning dynamics can become unstable, and convergence guarantees typically hold only in restricted settings such as two-player zero-sum or fully cooperative games. Moreover, when agents have heterogeneous and potentially conflicting preferences, it is unclear what system-level objective should guide learning. In this paper, we propose a new MARL pipeline called Near-Potential Policy Optimization (NePPO) for computing approximate Nash equilibria in mixed cooperative--competitive environments. The core idea is to learn a player-independent potential function such that the Nash equilibrium of a cooperative game with this potential as the common utility approximates a Nash equilibrium of the original game. To this end, we introduce a novel MARL objective such that minimizing this objective yields the best possible potential function candidate and consequently an approximate Nash equilibrium of the original game. We develop an algorithmic pipeline that minimizes this objective using zeroth-order gradient descent and returns an approximate Nash equilibrium policy. We empirically show the superior performance of this approach compared to popular baselines such as IPPO and MAPPO.

cs.LG↗

arXiv2026

Average Unfairness in Routing Games

We propose average unfairness as a new measure of fairness in routing games, defined as the ratio between the average latency and the minimum latency experienced by users. This measure is a natural complement to two existing unfairness notions: loaded unfairness, which compares maximum and minimum latencies of routes with positive flow, and user equilibrium (UE) unfairness, which compares maximum latency with the latency of a Nash equilibrium. We show that the worst-case values of all three unfairness measures coincide and are characterized by a steepness parameter intrinsic to the latency function class. We show that average unfairness is always no greater than loaded unfairness, and the two measures are equal only when the flow is fully fair. Besides that, we offer a complete comparison of the three unfairness measures, which, to the best of our knowledge, is the first theoretical analysis in this direction. Finally, we study the constrained system optimum (CSO) problem, where one seeks to minimize total latency subject to an upper bound on unfairness. We prove that, for the same tolerance level, the optimal flow under an average unfairness constraint achieves lower total latency than any flow satisfying a loaded unfairness constraint. We show that such improvement is always strict in parallel-link networks and establish sufficient conditions for general networks. We further illustrate the latter with numerical examples. Our results provide theoretical guarantees and valuable insights for evaluating fairness-efficiency tradeoffs in network routing.

cs.MA↗

arXiv2026

Game-to-Real Gap: Quantifying the Effect of Model Misspecification in Network Games

Game-theoretic models and solution concepts provide rigorous tools for predicting collective behavior in multi-agent systems. In practice, however, different agents may rely on different game-theoretic models to design their strategies. As a result, when these heterogeneous models interact, the realized outcome can deviate substantially from the outcome each agent expects based on its own local model. In this work, we introduce the game-to-real gap, a new metric that quantifies the impact of such model misspecification in multi-agent environments. The game-to-real gap is defined as the difference between the utility an agent actually obtains in the multi-agent environment (where other agents may have misspecified models) and the utility it expects under its own game model. Focusing on quadratic network games, we show that misspecifications in either (i) the external shock or (ii) the player interaction network can lead to arbitrarily large game-to-real gaps. We further develop novel network centrality measures that allow exact evaluation of this gap in quadratic network games. Our analysis reveals that standard network centrality measures fail to capture the effects of model misspecification, underscoring the need for new structural metrics that account for this limitation. Finally, through illustrative numerical experiments, we show that existing centrality measures in network games may provide a counterintuitive understanding of the impact of model misspecification.

cs.GT↗

arXiv2025

Opening the Sim-to-Real Door for Humanoid Pixel-to-Action Policy Transfer

Recent progress in GPU-accelerated, photorealistic simulation has opened a scalable data-generation path for robot learning, where massive physics and visual randomization allow policies to generalize beyond curated environments. Building on these advances, we develop a teacher-student-bootstrap learning framework for vision-based humanoid loco-manipulation, using articulated-object interaction as a representative high-difficulty benchmark. Our approach introduces a staged-reset exploration strategy that stabilizes long-horizon privileged-policy training, and a GRPO-based fine-tuning procedure that mitigates partial observability and improves closed-loop consistency in sim-to-real RL. Trained entirely on simulation data, the resulting policy achieves robust zero-shot performance across diverse door types and outperforms human teleoperators by up to 31.7% in task completion time under the same whole-body control stack. This represents the first humanoid sim-to-real policy capable of diverse articulated loco-manipulation using pure RGB perception.

cs.RO↗

arXiv2025

VIRAL: Visual Sim-to-Real at Scale for Humanoid Loco-Manipulation

A key barrier to the real-world deployment of humanoid robots is the lack of autonomous loco-manipulation skills. We introduce VIRAL, a visual sim-to-real framework that learns humanoid loco-manipulation entirely in simulation and deploys it zero-shot to real hardware. VIRAL follows a teacher-student design: a privileged RL teacher, operating on full state, learns long-horizon loco-manipulation using a delta action space and reference state initialization. A vision-based student policy is then distilled from the teacher via large-scale simulation with tiled rendering, trained with a mixture of online DAgger and behavior cloning. We find that compute scale is critical: scaling simulation to tens of GPUs (up to 64) makes both teacher and student training reliable, while low-compute regimes often fail. To bridge the sim-to-real gap, VIRAL combines large-scale visual domain randomization over lighting, materials, camera parameters, image quality, and sensor delays--with real-to-sim alignment of the dexterous hands and cameras. Deployed on a Unitree G1 humanoid, the resulting RGB-based policy performs continuous loco-manipulation for up to 54 cycles, generalizing to diverse spatial and appearance variations without any real-world fine-tuning, and approaching expert-level teleoperation performance. Extensive ablations dissect the key design choices required to make RGB-based humanoid loco-manipulation work in practice.

cs.RO↗

arXiv2025

Evader-Agnostic Team-Based Pursuit Strategies in Partially-Observable Environments

We consider a scenario where a team of two unmanned aerial vehicles (UAVs) pursue an evader UAV within an urban environment. Each agent has a limited view of their environment where buildings can occlude their field-of-view. Additionally, the pursuer team is agnostic about the evader in terms of its initial and final location, and the behavior of the evader. Consequently, the team needs to gather information by searching the environment and then track it to eventually intercept. To solve this multi-player, partially-observable, pursuit-evasion game, we develop a two-phase neuro-symbolic algorithm centered around the principle of bounded rationality. First, we devise an offline approach using deep reinforcement learning to progressively train adversarial policies for the pursuer team against fictitious evaders. This creates $k$-levels of rationality for each agent in preparation for the online phase. Then, we employ an online classification algorithm to determine a "best guess" of our current opponent from the set of iteratively-trained strategic agents and apply the best player response. Using this schema, we improved average performance when facing a random evader in our environment.

cs.MA↗

arXiv2025

Coordinated Autonomous Drones for Human-Centered Fire Evacuation in Partially Observable Urban Environments

Autonomous drone technology holds significant promise for enhancing search and rescue operations during evacuations by guiding humans toward safety and supporting broader emergency response efforts. However, their application in dynamic, real-time evacuation support remains limited. Existing models often overlook the psychological and emotional complexity of human behavior under extreme stress. In real-world fire scenarios, evacuees frequently deviate from designated safe routes due to panic and uncertainty. To address these challenges, this paper presents a multi-agent coordination framework in which autonomous Unmanned Aerial Vehicles (UAVs) assist human evacuees in real-time by locating, intercepting, and guiding them to safety under uncertain conditions. We model the problem as a Partially Observable Markov Decision Process (POMDP), where two heterogeneous UAV agents, a high-level rescuer (HLR) and a low-level rescuer (LLR), coordinate through shared observations and complementary capabilities. Human behavior is captured using an agent-based model grounded in empirical psychology, where panic dynamically affects decision-making and movement in response to environmental stimuli. The environment features stochastic fire spread, unknown evacuee locations, and limited visibility, requiring UAVs to plan over long horizons to search for humans and adapt in real-time. Our framework employs the Proximal Policy Optimization (PPO) algorithm with recurrent policies to enable robust decision-making in partially observable settings. Simulation results demonstrate that the UAV team can rapidly locate and intercept evacuees, significantly reducing the time required for them to reach safety compared to scenarios without UAV assistance.

cs.MA↗

arXiv2025

Learning to Grasp Anything by Playing with Random Toys

Robotic manipulation policies often struggle to generalize to novel objects, limiting their real-world utility. In contrast, cognitive science suggests that children develop generalizable dexterous manipulation skills by mastering a small set of simple toys and then applying that knowledge to more complex items. Inspired by this, we study if similar generalization capabilities can also be achieved by robots. Our results indicate robots can learn generalizable grasping using randomly assembled objects that are composed from just four shape primitives: spheres, cuboids, cylinders, and rings. We show that training on these "toys" enables robust generalization to real-world objects, yielding strong zero-shot performance. Crucially, we find the key to this generalization is an object-centric visual representation induced by our proposed detection pooling mechanism. Evaluated in both simulation and on physical robots, our model achieves a 67% real-world grasping success rate on the YCB dataset, outperforming state-of-the-art approaches that rely on substantially more in-domain data. We further study how zero-shot generalization performance scales by varying the number and diversity of training toys and the demonstrations per toy. We believe this work offers a promising path to scalable and generalizable learning in robotic manipulation. Demonstration videos, code, checkpoints and our dataset are available on our project page: https://lego-grasp.github.io/ .

cs.RO↗

arXiv2025

DreamControl: Human-Inspired Whole-Body Humanoid Control for Scene Interaction via Guided Diffusion

We introduce DreamControl, a novel methodology for learning autonomous whole-body humanoid skills. DreamControl leverages the strengths of diffusion models and Reinforcement Learning (RL): our core innovation is the use of a diffusion prior trained on human motion data, which subsequently guides an RL policy in simulation to complete specific tasks of interest (e.g., opening a drawer or picking up an object). We demonstrate that this human motion-informed prior allows RL to discover solutions unattainable by direct RL, and that diffusion models inherently promote natural looking motions, aiding in sim-to-real transfer. We validate DreamControl's effectiveness on a Unitree G1 robot across a diverse set of challenging tasks involving simultaneous lower and upper body control and object interaction. Project website at https://genrobo.github.io/DreamControl/

cs.RO↗

arXiv2025

Two-Stage Mechanism Design for Electric Vehicle Charging with Day-Ahead Reservations

We propose a general two-period model where electrical vehicles (EVs) can reserve charging sessions in the day-ahead market and swap them in the real-time market. Under the model, we explore several candidate mechanisms for running the two markets, compared using several normative properties such as incentive compatibility, efficiency, reservation awareness, and budget balance. Specifically, reservation awareness is the only property coupling the two markets and dictates that an EV will not get a lower utility by joining the real-time market. Focusing on the real-time market, we show that two variants of the classical Vickrey-Clarke-Groves (VCG) mechanism do not satisfy all the proposed properties; specifically, one is not reservation-aware, while the other is not budget-balanced. Moreover, we show that no mechanism satisfies some combinations of the properties. Then, we propose to use a posted-price mechanism to resolve the issue, which turns out to be the dynamic pricing mechanism adopted in many real-world systems. The proposed mechanism has no efficiency guarantee but satisfies all the other properties. To improve efficiency, we propose to use a VCG auction in the day-ahead market that guides the reserve prices in the real-time market. When EVs' valuations in the two markets are highly correlated, the proposed approach results in highly efficient outcomes.

eess.SY↗

arXiv2025

When are safety filters safe? On minimum phase conditions of control barrier functions

In emerging control applications involving multiple and complex tasks, safety filters are gaining prominence as a modular approach to enforcing safety constraints. Among various methods, control barrier functions (CBFs) are widely used for designing safety filters due to their simplicity, imposing a single linear constraint on the control input at each state. In this work, we focus on the internal dynamics of systems governed by CBF-constrained control laws. Our key observation is that, although CBFs guarantee safety by enforcing state constraints, they can inadvertently be "unsafe" by causing the internal state to diverge. We investigate the conditions under which the full system state, including the internal state, can remain bounded under a CBF-based safety filter. Drawing inspiration from the input-output linearization literature, where boundedness is ensured by minimum phase conditions, we propose a new set of CBF minimum phase conditions tailored to the structure imposed by the CBF constraint. A critical distinction from the original minimum phase conditions is that the internal dynamics in our setting is driven by a nonnegative virtual control input, which reflects the enforcement of the safety constraint. We include a range of numerical examples, including single-input, multi-input, linear, and nonlinear systems, validating both our analysis and the necessity of the proposed CBF minimum phase conditions.

eess.SY↗

arXiv2025

LeVERB: Humanoid Whole-Body Control with Latent Vision-Language Instruction

Vision-language-action (VLA) models have demonstrated strong semantic understanding and zero-shot generalization, yet most existing systems assume an accurate low-level controller with hand-crafted action "vocabulary" such as end-effector pose or root velocity. This assumption confines prior work to quasi-static tasks and precludes the agile, whole-body behaviors required by humanoid whole-body control (WBC) tasks. To capture this gap in the literature, we start by introducing the first sim-to-real-ready, vision-language, closed-loop benchmark for humanoid WBC, comprising over 150 tasks from 10 categories. We then propose LeVERB: Latent Vision-Language-Encoded Robot Behavior, a hierarchical latent instruction-following framework for humanoid vision-language WBC, the first of its kind. At the top level, a vision-language policy learns a latent action vocabulary from synthetically rendered kinematic demonstrations; at the low level, a reinforcement-learned WBC policy consumes these latent verbs to generate dynamics-level commands. In our benchmark, LeVERB can zero-shot attain a 80% success rate on simple visual navigation tasks, and 58.5% success rate overall, outperforming naive hierarchical whole-body VLA implementation by 7.8 times.

cs.RO↗

arXiv2025

LATTE-MV: Learning to Anticipate Table Tennis Hits from Monocular Videos

Physical agility is a necessary skill in competitive table tennis, but by no means sufficient. Champions excel in this fast-paced and highly dynamic environment by anticipating their opponent's intent - buying themselves the necessary time to react. In this work, we take one step towards designing such an anticipatory agent. Previous works have developed systems capable of real-time table tennis gameplay, though they often do not leverage anticipation. Among the works that forecast opponent actions, their approaches are limited by dataset size and variety. Our paper contributes (1) a scalable system for reconstructing monocular video of table tennis matches in 3D and (2) an uncertainty-aware controller that anticipates opponent actions. We demonstrate in simulation that our policy improves the ball return rate against high-speed hits from 49.9% to 59.0% as compared to a baseline non-anticipatory policy.

cs.CV↗