Behdad Chalaki

Harsh Goel, Mohammad Omama, Behdad Chalaki et al.

Multi-agent reinforcement learning (MARL) has achieved significant progress in large-scale traffic control, autonomous vehicles, and robotics. Drawing inspiration from biological systems where roles naturally emerge to enable coordination, role-based MARL methods have been proposed to enhance cooperation learning for complex tasks. However, existing methods exclusively derive roles from an agent's past experience during training, neglecting their influence on its future trajectories. This paper introduces a key insight: an agent's role should shape its future behavior to enable effective coordination. Hence, we propose Role Discovery and Diversity through Dynamics Models (R3DM), a novel role-based MARL framework that learns emergent roles by maximizing the mutual information between agents' roles, observed trajectories, and expected future behaviors. R3DM optimizes the proposed objective through contrastive learning on past trajectories to first derive intermediate roles that shape intrinsic rewards to promote diversity in future behaviors across different roles through a learned dynamics model. Benchmarking on SMAC and SMACv2 environments demonstrates that R3DM outperforms state-of-the-art MARL approaches, improving multi-agent coordination to increase win rates by up to 20%. The code is available at https://github.com/UTAustin-SwarmLab/R3DM.

Behdad Chalaki, Logan E. Beaver, Ben Remer et al.

In this article, we demonstrate a zero-shot transfer of an autonomous driving policy from simulation to University of Delaware's scaled smart city with adversarial multi-agent reinforcement learning, in which an adversary attempts to decrease the net reward by perturbing both the inputs and outputs of the autonomous vehicles during training. We train the autonomous vehicles to coordinate with each other while crossing a roundabout in the presence of an adversary in simulation. The adversarial policy successfully reproduces the simulated behavior and incidentally outperforms, in terms of travel time, both a human-driving baseline and adversary-free trained policies. Finally, we demonstrate that the addition of adversarial training considerably improves the performance \eat{stability and robustness} of the policies after transfer to the real world compared to Gaussian noise injection.

Lili Gao, Yanbo Xu, William Koch et al.

We introduce ScenarioControl, the first vision-language control mechanism for learned driving scenario generation. Given a text prompt or an input image, Scenario-Control synthesizes diverse, realistic 3D scenario rollouts - including map, 3D boxes of reactive actors over time, pedestrians, driving infrastructure, and ego camera observations. The method generates scenes in a vectorized latent space that represents road structure and dynamic agents jointly. To connect multimodal control with sparse vectorized scene elements, we propose a cross-global control mechanism that integrates crossattention with a lightweight global-context branch, enabling fine-grained control over road layout and traffic conditions while preserving realism. The method produces temporally consistent scenario rollouts from the perspectives different actors in the scene, supporting long-horizon continuation of driving scenarios. To facilitate training and evaluation, we release a dataset with text annotations aligned to vectorized map structures. Extensive experiments validate that the control adherence and fidelity of ScenarioControl compare favorable to all tested methods across all experiments. Project webpage: https://light.princeton.edu/ScenarioControl

Mohammad Omama, Po-han Li, Harsh Goel et al.

Mapping is crucial in robotics for localization and downstream decision-making. As robots are deployed in ever-broader settings, the maps they rely on continue to increase in size. However, storing these maps indefinitely (cold storage), transferring them across networks, or sending localization queries to cloud-hosted maps imposes prohibitive memory and bandwidth costs. We propose a text-enhanced compression framework that reduces both memory and bandwidth footprints while retaining high-fidelity localization. The key idea is to treat text as an alternative modality: one that can be losslessly compressed with large language models. We propose leveraging lightweight text descriptions combined with very small image feature vectors, which capture "complementary information" as a compact representation for the mapping task. Building on this, our novel technique, Similarity Space Replication (SSR), learns an adaptive image embedding in one shot that captures only the information "complementary" to the text descriptions. We validate our compression framework on multiple downstream localization tasks, including Visual Place Recognition as well as object-centric Monte Carlo localization in both indoor and outdoor settings. SSR achieves 2 times better compression than competing baselines on state-of-the-art datasets, including TokyoVal, Pittsburgh30k, Replica, and KITTI.

Muhan Lin, Shuyang Shi, Yue Guo et al. · cmu

The correct specification of reward models is a well-known challenge in reinforcement learning. Hand-crafted reward functions often lead to inefficient or suboptimal policies and may not be aligned with user values. Reinforcement learning from human feedback is a successful technique that can mitigate such issues, however, the collection of human feedback can be laborious. Recent works have solicited feedback from pre-trained large language models rather than humans to reduce or eliminate human effort, however, these approaches yield poor performance in the presence of hallucination and other errors. This paper studies the advantages and limitations of reinforcement learning from large language model feedback and proposes a simple yet effective method for soliciting and applying feedback as a potential-based shaping function. We theoretically show that inconsistent rankings, which approximate ranking errors, lead to uninformative rewards with our approach. Our method empirically improves convergence speed and policy returns over commonly used baselines even with significant ranking errors, and eliminates the need for complex post-processing of reward functions.

Vasanth Reddy Baddam, Behdad Chalaki, Vaishnav Tadiparthi et al.

In social robot navigation, traditional metrics like proxemics and behavior naturalness emphasize human comfort and adherence to social norms but often fail to capture an agent's autonomy and adaptability in dynamic environments. This paper introduces human empowerment, an information-theoretic concept that measures a human's ability to influence their future states and observe those changes, as a complementary metric for evaluating social compliance. This metric reveals how robot navigation policies can indirectly impact human empowerment. We present a framework that integrates human empowerment into the evaluation of social performance in navigation tasks. Through numerical simulations, we demonstrate that human empowerment as a metric not only aligns with intuitive social behavior, but also shows statistically significant differences across various robot navigation policies. These results provide a deeper understanding of how different policies affect social compliance, highlighting the potential of human empowerment as a complementary metric for future research in social navigation.

Sai Krishna Sumanth Nakka, Behdad Chalaki, Andreas Malikopoulos

The steady increase in the number of vehicles operating on the highways continues to exacerbate congestion, accidents, energy consumption, and greenhouse gas emissions. Emerging mobility systems, e.g., connected and automated vehicles (CAVs), have the potential to directly address these issues and improve transportation network efficiency and safety. In this paper, we consider a highway merging scenario and propose a framework for coordinating CAVs such that stop-and-go driving is eliminated. We use a decentralized form of the actor-critic approach to deep reinforcement learning$-$multi-agent deep deterministic policy gradient. We demonstrate the coordination of CAVs through numerical simulations and show that a smooth traffic flow is achieved by eliminating stop-and-go driving. Videos and plots of the simulation results can be found at this supplemental $\href{https://sites.google.com/view/ud-ids-lab/MADRL}{\text{site}}$.

Raymond M. Zayas, Logan E. Beaver, Behdad Chalaki et al.

The increasing demand for emerging mobility systems with connected and automated vehicles has imposed the necessity for quality testing environments to support their development. In this paper, we introduce a Unity-based virtual simulation environment for emerging mobility systems, called the Information and Decision Science Lab's Scaled Smart Digital City (IDS 3D City), intended to operate alongside its physical peer and its established control framework. By utilizing the Robot Operation System, AirSim, and Unity, we constructed a simulation environment capable of iteratively designing experiments significantly faster than it is possible in a physical testbed. This environment provides an intermediate step to validate the effectiveness of our control algorithms prior to their implementation in the physical testbed. The IDS 3D City also enables us to demonstrate that our control algorithms work independently of the underlying vehicle dynamics, as the vehicle dynamics introduced by AirSim operate at a different scale than our scaled smart city. Finally, we demonstrate the behavior of our digital environment by performing an experiment in both the virtual and physical environments and comparing their outputs.

Behdad Chalaki, Andreas A. Malikopoulos

Connected and automated vehicles (CAVs) can alleviate traffic congestion, air pollution, and improve safety. In this paper, we provide a decentralized coordination framework for CAVs at a signal-free intersection to minimize travel time and improve fuel efficiency. We employ a simple yet powerful reinforcement learning approach, an off-policy temporal difference learning called Q-learning, enhanced with a coordination mechanism to address this problem. Then, we integrate a first-in-first-out queuing policy to improve the performance of our system. We demonstrate the efficacy of our proposed approach through simulation and comparison with the classical optimal control method based on Pontryagin's minimum principle.

Behdad Chalaki, Logan E. Beaver, Andreas A. Malikopoulos

Roundabouts in conjunction with other traffic scenarios, e.g., intersections, merging roadways, speed reduction zones, can induce congestion in a transportation network due to driver responses to various disturbances. Research efforts have shown that smoothing traffic flow and eliminating stop-and-go driving can both improve fuel efficiency of the vehicles and the throughput of a roundabout. In this paper, we validate an optimal control framework developed earlier in a multi-lane roundabout scenario using the University of Delaware's scaled smart city (UDSSC). We first provide conditions where the solution is optimal. Then, we demonstrate the feasibility of the solution using experiments at UDSSC, and show that the optimal solution completely eliminates stop-and-go driving while preserving safety.

Kathy Jang, Eugene Vinitsky, Behdad Chalaki et al.

Using deep reinforcement learning, we train control policies for autonomous vehicles leading a platoon of vehicles onto a roundabout. Using Flow, a library for deep reinforcement learning in micro-simulators, we train two policies, one policy with noise injected into the state and action space and one without any injected noise. In simulation, the autonomous vehicle learns an emergent metering behavior for both policies in which it slows to allow for smoother merging. We then directly transfer this policy without any tuning to the University of Delaware Scaled Smart City (UDSSC), a 1:25 scale testbed for connected and automated vehicles. We characterize the performance of both policies on the scaled city. We show that the noise-free policy winds up crashing and only occasionally metering. However, the noise-injected policy consistently performs the metering behavior and remains collision-free, suggesting that the noise helps with the zero-shot policy transfer. Additionally, the transferred, noise-injected policy leads to a 5% reduction of average travel time and a reduction of 22% in maximum travel time in the UDSSC. Videos of the controllers can be found at https://sites.google.com/view/iccps-policy-transfer.