CyberSec Research

Model predictive control (MPC) combined with reduced-order template models has emerged as a powerful tool for trajectory optimization in dynamic legged locomotion. However, loco-manipulation tasks performed by legged robots introduce additional complexity, necessitating computationally efficient MPC algorithms capable of handling high-degree-of-freedom (DoF) models. This letter presents a computationally efficient nonlinear MPC (NMPC) framework tailored for loco-manipulation tasks of quadrupedal robots equipped with robotic manipulators whose dynamics are non-negligible relative to those of the quadruped. The proposed framework adopts a decomposition strategy that couples locomotion template models -- such as the single rigid body (SRB) model -- with a full-order dynamic model of the robotic manipulator for torque-level control. This decomposition enables efficient real-time solution of the NMPC problem in a receding horizon fashion at 60 Hz. The optimal state and input trajectories generated by the NMPC for locomotion are tracked by a low-level nonlinear whole-body controller (WBC) running at 500 Hz, while the optimal torque commands for the manipulator are directly applied. The layered control architecture is validated through extensive numerical simulations and hardware experiments on a 15-kg Unitree Go2 quadrupedal robot augmented with a 4.4-kg 4-DoF Kinova arm. Given that the Kinova arm dynamics are non-negligible relative to the Go2 base, the proposed NMPC framework demonstrates robust stability in performing diverse loco-manipulation tasks, effectively handling external disturbances, payload variations, and uneven terrain.

Navigation foundation models trained on massive webscale data enable agents to generalize across diverse environments and embodiments. However, these models trained solely on offline data, often lack the capacity to reason about the consequences of their actions or adapt through counterfactual understanding. They thus face significant limitations in the real-world urban navigation where interactive and safe behaviors, such as avoiding obstacles and moving pedestrians, are critical. To tackle these challenges, we introduce the Seeing-to-Experiencing framework to scale the capability of navigation foundation models with reinforcement learning. S2E combines the strengths of pre-training on videos and post-training through RL. It maintains the generalizability acquired from large-scale real-world videos while enhancing its interactivity through RL in simulation environments. Specifically, we introduce two innovations: an Anchor-Guided Distribution Matching strategy, which stabilizes learning and models diverse motion patterns through anchor-based supervision; and a Residual-Attention Module, which obtains reactive behaviors from simulation environments without erasing the model's pretrained knowledge. Moreover, we establish a comprehensive end-to-end evaluation benchmark, NavBench-GS, built on photorealistic 3DGS reconstructions of real-world scenes that incorporate physical interactions. It can systematically assess the generalizability and safety of navigation foundation models. Extensive experiments show that S2E mitigates the diminishing returns often seen when scaling with offline data alone. We perform a thorough analysis of the benefits of Reinforcement Learning compared to Supervised Fine-Tuning in the context of post-training for robot learning. Our findings emphasize the crucial role of integrating interactive online experiences to effectively scale foundation models in Robotics.

We present the first unified, modular, open-source 3DGS-based simulation framework for Real2Sim2Real robot learning. It features a holistic Real2Sim pipeline that synthesizes hyper-realistic geometry and appearance of complex real-world scenarios, paving the way for analyzing and bridging the Sim2Real gap. Powered by Gaussian Splatting and MuJoCo, Discoverse enables massively parallel simulation of multiple sensor modalities and accurate physics, with inclusive supports for existing 3D assets, robot models, and ROS plugins, empowering large-scale robot learning and complex robotic benchmarks. Through extensive experiments on imitation learning, Discoverse demonstrates state-of-the-art zero-shot Sim2Real transfer performance compared to existing simulators. For code and demos: https://air-discoverse.github.io/.

Retinal vein cannulation (RVC) is a minimally invasive microsurgical procedure for treating retinal vein occlusion (RVO), a leading cause of vision impairment. However, the small size and fragility of retinal veins, coupled with the need for high-precision, tremor-free needle manipulation, create significant technical challenges. These limitations highlight the need for robotic assistance to improve accuracy and stability. This study presents an automated robotic system with a top-down microscope and B-scan optical coherence tomography (OCT) imaging for precise depth sensing. Deep learning-based models enable real-time needle navigation, contact detection, and vein puncture recognition, using a chicken embryo model as a surrogate for human retinal veins. The system autonomously detects needle position and puncture events with 85% accuracy. The experiments demonstrate notable reductions in navigation and puncture times compared to manual methods. Our results demonstrate the potential of integrating advanced imaging and deep learning to automate microsurgical tasks, providing a pathway for safer and more reliable RVC procedures with enhanced precision and reproducibility.

Redundant manipulators are well understood to offer infinite joint configurations for achieving a desired end-effector pose. The multiplicity of inverse kinematics (IK) solutions allows for the simultaneous solving of auxiliary tasks like avoiding joint limits or obstacles. However, the most widely used IK solvers are numerical gradient-based iterative methods that inherently return a locally optimal solution. In this work, we explore the computation of self-motion manifolds (SMMs), which represent the set of all joint configurations that solve the inverse kinematics problem for redundant manipulators. Thus, SMMs are global IK solutions for redundant manipulators. We focus on task redundancies of dimensionality 1, introducing a novel ODE formulation for computing SMMs using standard explicit fixed-step ODE integrators. We also address the challenge of ``inducing'' redundancy in otherwise non-redundant manipulators assigned to tasks naturally described by one degree of freedom less than the non-redundant manipulator. Furthermore, recognizing that SMMs can consist of multiple disconnected components, we propose methods for searching for these separate SMM components. Our formulations and algorithms compute accurate SMM solutions without requiring additional IK refinement, and we extend our methods to prismatic joint systems -- an area not covered in current SMM literature. This manuscript presents the derivation of these methods and several examples that show how the methods work and their limitations.

One major recurring challenge in deploying manipulation robots is determining the optimal placement of manipulators to maximize performance. This challenge is exacerbated in complex, cluttered agricultural environments of high-value crops, such as flowers, fruits, and vegetables, that could greatly benefit from robotic systems tailored to their specific requirements. However, the design of such systems remains a challenging, intuition-driven process, limiting the affordability and adoption of robotics-based automation by domain experts like farmers. To address this challenge, we propose a four-part design optimization methodology for automating the development of task-specific robotic systems. This framework includes (a) a robot design model, (b) task and environment representations for simulation, (c) task-specific performance metrics, and (d) optimization algorithms for refining configurations. We demonstrate our framework by optimizing a dual-arm robotic system for pepper harvesting using two off-the-shelf redundant manipulators. To enhance performance, we introduce novel task metrics that leverage self-motion manifolds to characterize manipulator redundancy comprehensively. Our results show that our framework achieves simultaneous improvements in reachability success rates and improvements in dexterity. Specifically, our approach improves reachability success by at least 14\% over baseline methods and achieves over 30\% improvement in dexterity based on our task-specific metric.

Testing and evaluating automated driving systems (ADS) in interactions with vulnerable road users (VRUs), such as cyclists, are essential for improving the safety of VRUs, but often lack realism. This paper presents and validates a coupled in-the-loop test environment that integrates a Cyclist-in-the Loop test bench with a Vehicle-in-the-Loop test bench via a virtual environment (VE) developed in Unreal Engine 5. The setup enables closed-loop, bidirectional interaction between a real human cyclist and a real automated vehicle under safe and controllable conditions. The automated vehicle reacts to cyclist gestures via stimulated camera input, while the cyclist, riding a stationary bicycle, perceives and reacts to the vehicle in the VE in real time. Validation experiments are conducted using a real automated shuttle bus with a track-and-follow function, performing three test maneuvers - straight-line driving with stop, circular track driving, and double lane change - on a proving ground and in the coupled in-the-loop test environment. The performance is evaluated by comparing the resulting vehicle trajectories in both environments. Additionally, the introduced latencies of individual components in the test setup are measured. The results demonstrate the feasibility of the approach and highlight its strengths and limitations for realistic ADS evaluation.

Scientific testing techniques are essential for ensuring the safe operation of autonomous vehicles (AVs), with high-risk, highly interactive scenarios being a primary focus. To address the limitations of existing testing methods, such as their heavy reliance on high-quality test data, weak interaction capabilities, and low adversarial robustness, this paper proposes ExamPPO, an interactive adversarial testing framework that enables scenario-adaptive and intensity-controllable evaluation of autonomous vehicles. The framework models the Surrounding Vehicle (SV) as an intelligent examiner, equipped with a multi-head attention-enhanced policy network, enabling context-sensitive and sustained behavioral interventions. A scalar confrontation factor is introduced to modulate the intensity of adversarial behaviors, allowing continuous, fine-grained adjustment of test difficulty. Coupled with structured evaluation metrics, ExamPPO systematically probes AV's robustness across diverse scenarios and strategies. Extensive experiments across multiple scenarios and AV strategies demonstrate that ExamPPO can effectively modulate adversarial behavior, expose decision-making weaknesses in tested AVs, and generalize across heterogeneous environments, thereby offering a unified and reproducible solution for evaluating the safety and intelligence of autonomous decision-making systems.

Mobile manipulation is a critical capability for robots operating in diverse, real-world environments. However, manipulating deformable objects and materials remains a major challenge for existing robot learning algorithms. While various benchmarks have been proposed to evaluate manipulation strategies with rigid objects, there is still a notable lack of standardized benchmarks that address mobile manipulation tasks involving deformable objects. To address this gap, we introduce MoDeSuite, the first Mobile Manipulation Deformable Object task suite, designed specifically for robot learning. MoDeSuite consists of eight distinct mobile manipulation tasks covering both elastic objects and deformable objects, each presenting a unique challenge inspired by real-world robot applications. Success in these tasks requires effective collaboration between the robot's base and manipulator, as well as the ability to exploit the deformability of the objects. To evaluate and demonstrate the use of the proposed benchmark, we train two state-of-the-art reinforcement learning algorithms and two imitation learning algorithms, highlighting the difficulties encountered and showing their performance in simulation. Furthermore, we demonstrate the practical relevance of the suite by deploying the trained policies directly into the real world with the Spot robot, showcasing the potential for sim-to-real transfer. We expect that MoDeSuite will open a novel research domain in mobile manipulation involving deformable objects. Find more details, code, and videos at https://sites.google.com/view/modesuite/home.

A reliable communication network is essential for multiple UAVs operating within obstacle-cluttered environments, where limited communication due to obstructions often occurs. A common solution is to deploy intermediate UAVs to relay information via a multi-hop network, which introduces two challenges: (i) how to design the structure of multihop networks; and (ii) how to maintain connectivity during collaborative motion. To this end, this work first proposes an efficient constrained search method based on the minimumedge RRT? algorithm, to find a spanning-tree topology that requires a less number of UAVs for the deployment task. Then, to achieve this deployment, a distributed model predictive control strategy is proposed for the online motion coordination. It explicitly incorporates not only the inter-UAV and UAVobstacle distance constraints, but also the line-of-sight (LOS) connectivity constraint. These constraints are well-known to be nonlinear and often tackled by various approximations. In contrast, this work provides a theoretical guarantee that all agent trajectories are ensured to be collision-free with a teamwise LOS connectivity at all time. Numerous simulations are performed in 3D valley-like environments, while hardware experiments validate its dynamic adaptation when the deployment position changes online.

Visual Simultaneous Localization and Mapping (SLAM) plays a vital role in real-time localization for autonomous systems. However, traditional SLAM methods, which assume a static environment, often suffer from significant localization drift in dynamic scenarios. While recent advancements have improved SLAM performance in such environments, these systems still struggle with localization drift, particularly due to abrupt viewpoint changes and poorly characterized moving objects. In this paper, we propose a novel scene-object-based reliability assessment framework that comprehensively evaluates SLAM stability through both current frame quality metrics and scene changes relative to reliable reference frames. Furthermore, to tackle the lack of error correction mechanisms in existing systems when pose estimation becomes unreliable, we employ a pose refinement strategy that leverages information from reliable frames to optimize camera pose estimation, effectively mitigating the adverse effects of dynamic interference. Extensive experiments on the TUM RGB-D datasets demonstrate that our approach achieves substantial improvements in localization accuracy and system robustness under challenging dynamic scenarios.

The development of reinforcement learning (RL) algorithms has been largely driven by ambitious challenge tasks and benchmarks. Games have dominated RL benchmarks because they present relevant challenges, are inexpensive to run and easy to understand. While games such as Go and Atari have led to many breakthroughs, they often do not directly translate to real-world embodied applications. In recognising the need to diversify RL benchmarks and addressing complexities that arise in embodied interaction scenarios, we introduce Assistax: an open-source benchmark designed to address challenges arising in assistive robotics tasks. Assistax uses JAX's hardware acceleration for significant speed-ups for learning in physics-based simulations. In terms of open-loop wall-clock time, Assistax runs up to $370\times$ faster when vectorising training runs compared to CPU-based alternatives. Assistax conceptualises the interaction between an assistive robot and an active human patient using multi-agent RL to train a population of diverse partner agents against which an embodied robotic agent's zero-shot coordination capabilities can be tested. Extensive evaluation and hyperparameter tuning for popular continuous control RL and MARL algorithms provide reliable baselines and establish Assistax as a practical benchmark for advancing RL research for assistive robotics. The code is available at: https://github.com/assistive-autonomy/assistax.

With the rapid advancement of autonomous driving technology, vehicle-to-everything (V2X) communication has emerged as a key enabler for extending perception range and enhancing driving safety by providing visibility beyond the line of sight. However, integrating multi-source sensor data from both ego-vehicles and infrastructure under real-world constraints, such as limited communication bandwidth and dynamic environments, presents significant technical challenges. To facilitate research in this area, we organized the End-to-End Autonomous Driving through V2X Cooperation Challenge, which features two tracks: cooperative temporal perception and cooperative end-to-end planning. Built on the UniV2X framework and the V2X-Seq-SPD dataset, the challenge attracted participation from over 30 teams worldwide and established a unified benchmark for evaluating cooperative driving systems. This paper describes the design and outcomes of the challenge, highlights key research problems including bandwidth-aware fusion, robust multi-agent planning, and heterogeneous sensor integration, and analyzes emerging technical trends among top-performing solutions. By addressing practical constraints in communication and data fusion, the challenge contributes to the development of scalable and reliable V2X-cooperative autonomous driving systems.

Multi-robot SLAM aims at localizing and building a map with multiple robots, interacting with each other. In the work described in this article, we analyze the pipeline of a decentralized LiDAR SLAM system to study the current limitations of the state of the art, and we discover a significant source of failures, i.e., that the loop detection is the source of too many false positives. We therefore develop and propose a new heuristic to overcome these limitations. The environment taken as reference in this work is the highly challenging case of underground tunnels. We also highlight potential new research areas still under-explored.

Modeling and evaluation of automated vehicles (AVs) in mixed-autonomy traffic is essential prior to their safe and efficient deployment. This is especially important at urban junctions where complex multi-agent interactions occur. Current approaches for modeling vehicular maneuvers and interactions at urban junctions have limitations in formulating non-cooperative interactions and vehicle dynamics within a unified mathematical framework. Previous studies either assume predefined paths or rely on cooperation and central controllability, limiting their realism and applicability in mixed-autonomy traffic. This paper addresses these limitations by proposing a modeling framework for trajectory planning and decentralized vehicular control at urban junctions. The framework employs a bi-level structure where the upper level generates kinematically feasible reference trajectories using an efficient graph search algorithm with a custom heuristic function, while the lower level employs a predictive controller for trajectory tracking and optimization. Unlike existing approaches, our framework does not require central controllability or knowledge sharing among vehicles. The vehicle kinematics are explicitly incorporated at both levels, and acceleration and steering angle are used as control variables. This intuitive formulation facilitates analysis of traffic efficiency, environmental impacts, and motion comfort. The framework's decentralized structure accommodates operational and stochastic elements, such as vehicles' detection range, perception uncertainties, and reaction delay, making the model suitable for safety analysis. Numerical and simulation experiments across diverse scenarios demonstrate the framework's capability in modeling accurate and realistic vehicular maneuvers and interactions at various urban junctions, including unsignalized intersections and roundabouts.

Robotic task planning in real-world environments requires reasoning over implicit constraints from language and vision. While LLMs and VLMs offer strong priors, they struggle with long-horizon structure and symbolic grounding. Existing methods that combine LLMs with symbolic planning often rely on handcrafted or narrow domains, limiting generalization. We propose UniDomain, a framework that pre-trains a PDDL domain from robot manipulation demonstrations and applies it for online robotic task planning. It extracts atomic domains from 12,393 manipulation videos to form a unified domain with 3137 operators, 2875 predicates, and 16481 causal edges. Given a target class of tasks, it retrieves relevant atomics from the unified domain and systematically fuses them into high-quality meta-domains to support compositional generalization in planning. Experiments on diverse real-world tasks show that UniDomain solves complex, unseen tasks in a zero-shot manner, achieving up to 58% higher task success and 160% improvement in plan optimality over state-of-the-art LLM and LLM-PDDL baselines.

Human demonstration data is often ambiguous and incomplete, motivating imitation learning approaches that also exhibit reliable planning behavior. A common paradigm to perform planning-from-demonstration involves learning a reward function via Inverse Reinforcement Learning (IRL) then deploying this reward via Model Predictive Control (MPC). Towards unifying these methods, we derive a replacement of the policy in IRL with a planning-based agent. With connections to Adversarial Imitation Learning, this formulation enables end-to-end interactive learning of planners from observation-only demonstrations. In addition to benefits in interpretability, complexity, and safety, we study and observe significant improvements on sample efficiency, out-of-distribution generalization, and robustness. The study includes evaluations in both simulated control benchmarks and real-world navigation experiments using few-to-single observation-only demonstrations.

This work focuses on multi-floor indoor exploration, which remains an open area of research. Compared to traditional methods, recent learning-based explorers have demonstrated significant potential due to their robust environmental learning and modeling capabilities, but most are restricted to 2D environments. In this paper, we proposed a learning-integrated topological explorer, LITE, for multi-floor indoor environments. LITE decomposes the environment into a floor-stair topology, enabling seamless integration of learning or non-learning-based 2D exploration methods for 3D exploration. As we incrementally build floor-stair topology in exploration using YOLO11-based instance segmentation model, the agent can transition between floors through a finite state machine. Additionally, we implement an attention-based 2D exploration policy that utilizes an attention mechanism to capture spatial dependencies between different regions, thereby determining the next global goal for more efficient exploration. Extensive comparison and ablation studies conducted on the HM3D and MP3D datasets demonstrate that our proposed 2D exploration policy significantly outperforms all baseline explorers in terms of exploration efficiency. Furthermore, experiments in several 3D multi-floor environments indicate that our framework is compatible with various 2D exploration methods, facilitating effective multi-floor indoor exploration. Finally, we validate our method in the real world with a quadruped robot, highlighting its strong generalization capabilities.