News & Research

Social robot navigation presents a significant and complex challenge within indoor environments characterized by numerous obstacles and dynamic changes. Reinforcement learning (RL) serves as a potent approach for addressing this issue. However, the primary attributes of such environments—their diversity and complexity—pose substantial challenges to the generalization capabilities of RL-based social robot navigation systems. In this article, we propose a meta-reinforcement learning framework for social robot navigation that empowers robots to adapt to a variety of environments. Furthermore, we introduce the concept of utilizing social norms to guide social navigation for the first time. We employ field modeling to represent interactions among the environment, pedestrians, and robots. Historical pedestrian motion data are utilized as training input for physical information neural networks, which generate pedestrian movement vector fields within the environment. We then incorporate the similarity between the robot’s predicted motion state and its actual motion state into the reward function, guiding the robot’s learning process in alignment with social norms. Simulation results demonstrate that our proposed method enhances the robot’s social navigation capabilities.

Real-time collision avoidance is a critical challenge in visual servoing control for safe robot manipulation. This article presents a model-free framework that integrates reinforcement learning (RL) with real-time collision detection (RCD), enabling secure and smooth visual servoing in dynamic environments. The reward function in the RL framework is carefully designed to simultaneously minimize tracking errors, promote motion smoothness, and enforce collision avoidance. Beyond policy optimization through RL, a two-step real-time trajectory blending (RTB) scheme, consisting of a Butterworth filter (BF) and polynomial interpolation, is developed to further refine the generated trajectories and ensure smooth joint velocity profiles. Furthermore, the soft actor-critic (SAC) algorithm is employed for policy learning, and the performance is compared with two widely used methods, namely proximal policy optimization (PPO) and twin delayed deep deterministic policy gradient (TD3). Experimental results on a 6-degree-of-freedom (6-DoF) robot manipulator demonstrate the effectiveness of the proposed approach in achieving safe and smooth visual servoing in real time.

Quadruped robots are used for primary searches during the early stages of indoor fires. A typical primary search involves quickly and thoroughly looking for victims under hazardous conditions and monitoring flammable materials. However, situational awareness in complex indoor environments and rapid stair climbing across different staircases remain the main challenges for robot-assisted primary searches. In this project, we designed a two-stage end-to-end deep reinforcement learning (RL) approach to optimize both navigation and locomotion. In the first stage, the quadrupeds, Unitree Go2, were trained to climb stairs in Isaac Lab's pyramid-stair terrain. In the second stage, the quadrupeds were trained to climb various realistic indoor staircases in the Isaac Lab engine, with the learned policy transferred from the previous stage. These indoor staircases are straight, L-shaped, and spiral, to support climbing tasks in complex environments. This project explores how to balance navigation and locomotion and how end-to-end RL methods can enable quadrupeds to adapt to different stair shapes. Our main contributions are: (1) A two-stage end-to-end RL framework that transfers stair-climbing skills from abstract pyramid terrain to realistic indoor stair topologies. (2) A centerline-based navigation formulation that enables unified learning of navigation and locomotion without hierarchical planning. (3) Demonstration of policy generalization across diverse staircases using only local height-map perception. (4) An empirical analysis of success, efficiency, and failure modes under increasing stair difficulty.

Diffusion-based reinforcement learning dem effectiveness of the proposed algorithm is validated through simulations on the MuJoCo platform and experiments on a physical robotic platform. The relevant code for this study is publicly available on GitHub: https://github.com/zhn-ya/diffusion-RL.

We present a coverage framework that integrates Hilbert space-filling priors into decentralized multi-robot learning and execution. We augment DQN and PPO with Hilbert-based spatial indices to structure exploration and reduce redundancy in sparse-reward environments, and we evaluate scalability in multi-robot grid coverage. We further describe a waypoint interface that converts Hilbert orderings into curvature-bounded, time-parameterized SE(2) trajectories (planar (x, y, {\theta})), enabling onboard feasibility on resource-constrained robots. Experiments show improvements in coverage efficiency, redundancy, and convergence speed over DQN/PPO baselines. In addition, we validate the approach on a Boston Dynamics Spot legged robot, executing the generated trajectories in indoor environments and observing reliable coverage with low redundancy. These results indicate that geometric priors improve autonomy and scalability for swarm and legged robotics.

Background: Coordinating differential-drive mobile robots for landmark coverage is challenging due to non-holonomic dynamics, clutter, and sparse rewards. Standard multi-agent RL pipelines often show unstable learning and inconsistent completion in this setting. Methodology: We adopt a centralized-training, decentralized-execution actor-critic without inter-agent commu nication. Our replay-centric design combines a tagged buffer that up-samples goal-reaching transitions and an offline replay initialization that seeds early learning with curated trajectories. Dynamic task assignment uses the Hungarian algorithm during training and evaluation, and we benchmark against uniform replay and established variants. Results: In a cluttered six-robot arena, the approach improves training stability relative to uniform replay. Convergence is faster and requires fewer updates to reach consistent success. Coverage efficiency increases as landmarks are reached earlier across runs. Collisions per episode decrease without adding communication or architectural changes. Multi seed evaluations show gains that persist with narrow confidence intervals. Train-evaluation gaps shrink on unseen maps, indicating improved generalization. Ablations attribute complementary ben efits to the tagged and offline components. Performance remains competitive with prioritized and hindsight replay baselines under matched budgets. Computational overhead is small because sam pling logic changes while network sizes do not. Conclusions: Focus ing on replay design substantially stabilizes multi-agent learning for differential-drive coordination. The pipeline integrates cleanly with standard CTDE implementations and supports practical deployment in coverage tasks.

Recent advances in learned control, large-scale simulation, and generative models have accelerated progress toward general-purpose robotic controllers, yet the field still lacks platforms suitable for safe, expressive, long-term deployment in human environments. Most existing humanoids are either closed industrial systems or academic prototypes that are difficult to deploy and operate around people, limiting progress in robotics. We introduce Sprout, a developer platform designed to address these limitations through an emphasis on safety, expressivity, and developer accessibility. Sprout adopts a lightweight form factor with compliant control, limited joint torques, and soft exteriors to support safe operation in shared human spaces. The platform integrates whole-body control, manipulation with integrated grippers, and virtual-reality-based teleoperation within a unified hardware-software stack. An expressive head further enables social interaction -- a domain that remains underexplored on most utilitarian humanoids. By lowering physical and technical barriers to deployment, Sprout expands access to capable humanoid platforms and provides a practical basis for developing embodied intelligence in real human environments.

This paper presents an Agentic AI-based framework for automated safety analysis of humanoid and service robots using their CAD designs. The framework integrates multiple specialized AI agents and tools that collaboratively import and analyze CAD models to detect and segment hazardous features such as sharp edges, gaps, holes, pinch points, and exposed moving parts (e.g., limbs, wheels, gears, and motors). By combining vision-language models (VLMs), geometric analysis, and semantic segmentation, the system evaluates compliance with international safety standards. Identified semantic features such as gaps, holes, accessible areas, sharp edges, slots, and pinch points are correlated with ISO defined safety thresholds, enabling the AI agents to assess compliance, estimate the likelihood of injury or entrapment, and determine whether the robot's CAD-based form factors meet safety requirements. This work defines key Agentic AI use cases, outlines the design and workflow of collaborating agents and tools, and proposes an integrated design-support system for early-stage robot development, enabling proactive hazard analysis & risk assessment throughout product design life cycle.

Aiming at dynamic jumping control problem of humanoid bipedal robots, parametric mechanism design, kinematic/dynamic model, numerical solution algorithm of dynamic parameters, motion simulation and prototype experiment of the underactuated takeoff process of humanoid bipedal robots are proposed. The underactuated takeoff process of the robot is divided into the stance phase before takeoff (dominated by the driving joint) and the underactuated takeoff phase (passively rotating around tiptoe). The kinematic model of underactuated takeoff process is established by using the D 'Alembert principle, and the kinematic criterion for the robot entering underactuated phase is derived. The motion critical conditions during takeoff process are analyzed, and trajectory planning method of the total centroid based on variable quartic polynomial interpolation is proposed. The dynamic constraints of toe rotation in underactuated phase are constructed. The dynamic analytical model of underactuated takeoff is established by using the Lagrange equation, and the closed-loop solution of joint spatial trajectories are achieved through inverse kinematic. Finally, a series of motion simulation and prototype experiment are carried out to verify effectiveness of the proposed mechanism design, kinematic/dynamic modeling, trajectory planning algorithm and motion control strategy.

With the advancement of humanoid robotics, human–robot collaboration has emerged as a prominent research focus. Ensuring the safety of both humanoid robots and humans remains a critical challenge. In this paper, we address conflict resolutions at the planning level and propose a safety-oriented motion planning (SOMP) algorithm for a wheeled humanoid robot operating in environments with unknown human motions. In the proposed SOMP algorithm, we employ Monte Carlo simulations to predict trajectories of stochastically moving humans and formulate both hard and soft constraints. A dynamic-quadrant stochastic sampling policy, integrated with a rapidly exploring random tree method, is proposed to generate diverse initial paths. Building upon this, we develop a constraint-fusion mechanism that combines hard constraints for safety guarantees and soft constraints for path optimization, thereby effectively resolving potential conflicts between wheeled humanoid robots and stochastically moving humans. We evaluate the proposed algorithm under different configurations of conflict numbers, task success rates, and path rewards. The proposed method outperforms A*, RRT, and MDP in terms of conflict numbers (−77.8%, −76.6%, and −71.4%) and task success rates (+168.0%, +109.4%, and +91.4%). Our simulation results prove the efficiency and robustness of our algorithm in safe motion planning with stochastically moving humans.