News & Research

Labor shortages and aging populations are straining hospital systems globally, prompting interest in humanoid robots as a viable technology solution for healthcare support. This paper presents a proof-of-concept deployment of a Unitree G1 humanoid robot using advanced SLAM with multi-modal sensors (LiDAR, depth cameras) for reliable navigation in dynamic hospital-like environments. Unlike conventional approaches that model humans solely as obstacles to avoid, our system implements human-aware navigation. Respecting proxemics, social context, and malleable spatial boundaries to treat people as collaborative partners, enabling safe, legible co-navigation in crowded wards and corridors. By demonstrating autonomous, purposeful movement for logistics and patient-support tasks, this research illustrates a pathway to augment healthcare delivery, worker support, and patient care outcomes through adaptable, human-centered robotics.

This paper deals with a problem of multitask agents' navigation in urban environment with obstacles using different algorithmic approaches: neural network models with a system of rewards and penalties, neural networks oriented on distance minimization, as well as models that use direct movement vectors and rectangular route structures. The novelty of the presented method is the creation of a hybrid algorithm that incorporates strict rules for movement in an initially unknown environment and, when obstacles appear, switches to a neural network for obstacle avoidance. This approach combines deterministic movement, which can be adapted to artificial intelligence for safe and reliable navigation in uncharted territory. Experiments showed that thirty neural argents trained purely using rewards and penalties in movement exhibits slow learning and cyclical movement as clusters, which is due to the fact that they optimize the current state without global planning. Models trained solely on distance optimization perform slightly better, but still create clusters of movement toward the goal. Meanwhile, the use of strict linearity rules and rectangular routes allows for effective obstacle avoidance and consistent target achievement with less time under conditions of limited graph information. The results show that rule-based learning yields better results in agent movement than fully autonomous agents. Consequently, the combined path generation method yields the best results in agent navigation along a known path.

With the rapid development of Internet technology and multimedia applications, and the continuous expansion of interactive scenes of humanoid robot education, the number of digital resources of ethnic music has reached an unprecedented scale and continues to grow. This study mainly focuses on the construction of a national music retrieval system that integrates multimodal attention mechanisms and its application implementation in humanoid robot education interaction. Firstly, a method for automatic annotation of ethnic music adapted to multimodal data was studied to meet the human-computer interaction needs of robot education interaction. Secondly, a label conditional random field music automatic annotation method is proposed, and a multi-modal attention mechanism deep neural network model for ethnic music annotation is constructed to enhance the accurate matching of music features and retrieval requirements in human-computer interaction. Finally, integrate the constructed ethnic music retrieval system into the humanoid robot education interactive platform, conduct application analysis based on actual educational interactive scenarios, verify the effectiveness of the system in the human-computer interaction process, and complete performance evaluation. The results indicate that in the Glu module, Glu blocks perform better in ethnic music annotation tasks. The annotation results of various indicators in music hierarchy sequence modeling are superior to traditional models, ensuring the accuracy of ethnic music annotation in human-computer interaction scenarios. Compared with other algorithms, the AUC label score of this system is the highest, reaching 0.913. It can more efficiently model the mapping relationship between multimodal music features input during human-computer interaction and text retrieval requirements. It shows better performance in all evaluation indicators and can effectively support ethnic music retrieval services in humanoid robot education interaction, enhancing the immersion and fun of educational interaction.

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.

Efficient representation and querying of railway networks are crucial for autonomous railway systems and digital infrastructure management. This paper introduces VEctorial Railway NEtwork (VERNE), an interpretable data structure and algorithm that integrates vector-based spatial partitioning with a railway-specific topological framework to enhance network representation and navigation. VERNE is designed to optimize query efficiency, reduce memory footprint, and ensure scalability for real-time applications. Its internal mechanism results from a comparative performance analysis between a $k$-d tree, an STRtree and two custom algorithms, highlighting trade-offs in computational efficiency and memory overhead. The proposed approach is validated using datasets from both the French and Swedish railway networks, demonstrating its effectiveness in real-world scenarios. The results indicate that VERNE provides a robust and scalable solution for railway infrastructure modeling, offering improvements in localization speed and computational efficiency. Another advantage is that it inherently manipulates atomic elements which can contain any information relevant to directly perform navigation onboard an autonomous robot. This work contributes to the advancement of railway digitalization by providing a structured methodology for spatial data processing in autonomous railway systems.

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.

Humanoid robots must navigate, decide, and schedule efficiently to boost automated supply chain efficiency. Traditional rule-based techniques fail in dynamic situations, especially with task dependencies. A unique Puma Optimizer-mutated Twin-Stage Adaptive Twin-Delayed Deep Deterministic Policy Gradient (PO-TSATD3) method is used in this deep reinforcement learning system. Training datasets imitate real-world logistics situations with dynamic impediments, many robots, and varying workloads. Data preparation cleans and normalizes for quality. The Puma Optimizer optimises convergence and operating efficiency, while the PO-TSATD3 framework improves navigation and scheduling adaptive learning. Python simulations show considerable gains in navigation accuracy, collision reduction, and schedule optimization over conventional methods. The model's outstanding performance metrics proved its scalability and durability in complicated situations. This research validates the application of deep reinforcement learning, augmented by PO-TSATD3, as a powerful solution for intelligent humanoid robot operations in future supply chain systems.

Autonomous navigation in cluttered, unknown environments requires agents to balance local obstacle avoidance with global path optimality. This paper presents a novel hybrid approach for path planning in such environments by integrating Spiking Neural Networks, Artificial Potential Fields, and RRT* for episodic path optimization based on spatial memory. The proposed approach can be viewed as a lifelong learning paradigm where the intelligent agent creates a cognitive map based on the Spiking Neural Networks and continuously updates its spatial memory during episodes of navigation based on the Artificial Potential Fields approach. The spatial memory obtained from the Artificial Potential Fields approach can then be integrated into the RRT* path planning approach for path optimization during subsequent episodes of navigation, avoiding certain areas where obstacles are more probable based on the spatial memory. The performance of the proposed approach was evaluated on a grid environment with various obstacles, and the results show rapid convergence within six episodes with a success rate of 100% and a reduction in path loss of 89.8% for the normalized path length. The performance of the proposed approach was compared with the Artificial Potential Fields and RRT* path planning approaches, and the results show the proposed approach can significantly reduce the planning time and node expansion for path planning. The proposed approach can therefore be viewed as a novel approach for path planning based on spatial memory, which can be used for efficient path planning in unknown and dynamic environments.

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.

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.

We showed recently that a novel prosthesis control, based on natural arm movement reconstructed from an Artificial Neural Network (ANN) that receives movement goal and stump motion as inputs, enables close-to-natural reaching control in Virtual Reality (VR). Here, we applied and tested this control in the physical world using first-person teleoperation of Reachy2, an open-source humanoid robot made by Pollen Robotics that ranked 2nd at ANA Avatar XPRIZE. Direct teleoperation reproducing all arm joints of non-disabled participants (n=14) elicited a 100% success rate for grasping objects at various locations, with best possible usability scores and small workload scores. High success rates (>92%) were also obtained when all distal joints from the elbow onward were operated with the novel prosthesis control, using a movement goal identified through gaze-guided computer vision. While usability and workload scores were slightly degraded when non-disabled participants used the prosthesis control as compared to direct teleoperation, both were scored similarly well by a sample from our target population (n=8 participants with transhumeral limb loss). This platform and control schemes offer broad perspectives for prosthesis control, but also for teleoperation robotics and human-robot interactions.

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.