Safety and productivity are two critical aspects of human–robot collaborative (HRC) environments. This paper proposes ranked behavior cloning (RBC), an imitation learning framework that enhances conventional behavior cloning by incorporating a ranking mechanism based on cooperative speed-up (CSU) and jerk-based smoothness. The proposed framework addresses safety at two distinct levels: (1) learned safety-awareness, where the ranking objective penalizes high-jerk, low-productivity demonstrations so the trained policy inherently produces smoother, more cooperative trajectories; and (2) externally enforced safety, implemented as a zone-based Speed and Separation Monitoring (SSM) layer that dynamically adjusts robot speed based on human–robot proximity, inspired by the SSM principles defined in ISO 15066. The proposed RBC model is evaluated on a real-world UR5 collaborative platform across five manipulation tasks, benchmarked against standard behavior cloning (BC) and Maximum Entropy Inverse Reinforcement Learning (MaxEnt-IRL). RBC achieves a positional accuracy of 92.6
Reinforcement learning (RL) has enabled agile legged locomotion; however, transferring these results to a new robot often requires extensive reward redesign and coefficient retuning as the morphology, scale, and actuation constraints change. This study evaluates a gait-parameterized RL formulation for cross-platform locomotion learning without reward retuning. We keep the reward terms and coefficients, network architecture, and PPO hyperparameters fixed and adjust only the platform interface via observation/action mappings and low-level actuation settings. With this setup, we trained separate policies in the simulation on three small quadrupeds (Go1, ANYmal C, and Mini Cheetah), a humanoid (Atlas), and a large-scale hydraulic quadruped (BeTheX-Q), and evaluated their performance across multiple terrains. The results showed stable locomotion across large variations in morphology and physical scale. The learned gait-intent outputs further exhibited systematic scaling and correlation trends with speed, posture, actuation effort, and mechanical cost of transport, thereby providing an interpretable view of embodiment-dependent behaviors under a fixed learning formulation. Finally, we extend the framework to a wheeled-legged hybrid configuration of BeTheX-Q and demonstrate wheel–leg coupled locomotion in both simulation and hardware.
In this paper, we investigate the Minimum Obstacle Displacement Planning problem from a robot motion planning perspective. The problem involves determining a feasible path to a goal location by displacing movable obstacles when no collision-free path initially exists. We show that this problem is computationally challenging and, in particular, NP-hard when obstacles are modeled as polygons in the plane. Besides an exact formulation of the minimum obstacle displacement problem generalizing other problems in the literature, and the associated optimal solution, this paper proposes an approximate solution that is less intensive from a computational standpoint, and differs from the optimal solution by a fraction of the optimal cost, being able to trade-off between path length and amount of obstacle displacements.
Path planning for manipulators in cluttered environments remains challenging because conventional RRT* often suffers from excessive random sampling, invalid extensions, and paths with sharp local turns. This paper proposes IBPAS-RRT*, an adaptive bidirectional RRT* algorithm with joint-space potential field guidance. Instead of relying on task-space planning followed by repeated inverse kinematics, the proposed method incorporates obstacle and goal information into joint-space tree expansion, thereby avoiding repeated inverse kinematics calls during planning. An adaptive goal-bias mechanism is introduced to regulate the sampling preference according to collision-related extension feedback. The potential field component, random exploration component, and expansion step are also coordinated to improve local search adaptability in cluttered regions. After path generation, pruning and B-spline smoothing are applied to reduce redundant waypoints and improve path continuity. Experiments in seven 2D, 3D, and joint-space scenarios show that IBPAS-RRT* reduces the iteration number and planning time by up to 54.46
This study proposes a multimodal robotic task planning framework based on large language models (LLMs) that utilizes both visual and force-derived physical feedback. The system integrates multimodal inputs, including camera images and force/torque (F/T) sensor data, to interpret the visual and physical properties of objects and generate a task sequence to execute given commands. The integration of vision and force data allows the system to compensate for the limitations of each modality, leading to more reliable decision-making. The framework consists of three components: Extractor, Planner, and Sub-planner. According to their assigned roles, each agent automatically translates high-level commands given in natural language into executable robot motion plans, enabling the robot to perform the required sequence of actions. The system combines visual and force data to handle tasks that are difficult or infeasible with a single modality. Furthermore, it supports sequential and condition-based task planning through the collaboration of multiple LLM agents. Experimental results in various task scenarios show that the proposed framework consistently improves overall task success rates compared with unimodal settings with different LLMs and achieves a higher success rate compared to using only visual or force data.
