KANGAROO
Configurable bipedal humanoid platform from PAL Robotics for locomotion and reinforcement-learning research, offered with 4-, 5- or 7-DoF arms and ROS 2 software.
Machines, companies, technologies, industries, applications, components, people, research, rules, glossary terms and attributed news, ranked in one list.
56 results for “reinforcement learning research”
Configurable bipedal humanoid platform from PAL Robotics for locomotion and reinforcement-learning research, offered with 4-, 5- or 7-DoF arms and ROS 2 software.
arXiv:2611.08814v1 Announce Type: new Abstract: Geometric assembly requires fine spatiotemporal coordination to place parts without causing collisions or deadlocks under…
arXiv:2610.41348v1 Announce Type: new Abstract: Robots deployed in unstructured environments are susceptible to actuator-authority degradation, which reduces the…
arXiv:2610.31812v1 Announce Type: new Abstract: Preference-based reinforcement learning (PbRL) has emerged as a promising paradigm for learning robot behaviors aligned…
arXiv:2610.31829v1 Announce Type: new Abstract: Real-world reinforcement learning has enabled contact-rich manipulation tasks to be trained directly on physical robots…
arXiv:2611.19616v1 Announce Type: new Abstract: Reinforcement learning (RL) enables robots to acquire complex skills, but standard RL formulations do not guarantee that…
arXiv:2611.25232v1 Announce Type: new Abstract: Ferrofluids are liquids containing magnetic nanoparticles, enabling shape control through external magnetic fields. This…
arXiv:2610.32143v1 Announce Type: new Abstract: Robot-assisted dressing is particularly challenging due to narrow openings and articulated body parts. Even small errors…
arXiv:2611.19324v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models have demonstrated strong capabilities in multi-task robotic manipulation. However…
arXiv:2610.33480v1 Announce Type: new Abstract: Dexterous manipulation requires high-dimensional control policies to accomplish complex contact-rich tasks while…
arXiv:2611.19318v1 Announce Type: new Abstract: We address tracking a moving object on a conveyor with a Franka Emika Panda robot arm, despite an imprecise digital twin…
Unitree's compact bipedal humanoid, about 1.32 m tall and 35 kg, sold in a base version and an EDU version that adds dexterous hands, extra waist/wrist joints and…
Pieter Abbeel is a UC Berkeley professor working at the intersection of robotics and machine learning.
A lower-body bipedal platform from EngineAI for research and education, released in July 2024 with open-source software access. It has 12 degrees of freedom (six per…
arXiv:2611.06196v1 Announce Type: new Abstract: Ensuring safety in advanced robot control methods such as Model Predictive Control (MPC) and Reinforcement Learning…
arXiv:2610.32093v1 Announce Type: new Abstract: Behavior cloning for robotic manipulation is highly sensitive to the coverage of its training data, yet collecting…
arXiv:2611.25320v1 Announce Type: cross Abstract: Residual reinforcement learning corrects an imitation-trained base policy using a limited budget of environment…
arXiv:2611.08857v1 Announce Type: new Abstract: Learning locomotion directly on a humanoid robot is a promising direction to avoid the sim-to-real gap. However…
A small 3D-printed, servo-driven open-source humanoid for sim-to-real and reinforcement-learning experiments, with a published bill of materials starting around $350…
arXiv:2611.08872v1 Announce Type: new Abstract: In recent years, massively parallelized rigid-body simulation has played a central role in robotics research, enabling…
arXiv:2605.05236v3 Announce Type: replace Abstract: In the field of precision manufacturing in complex constrained environments, the role of soft robots is increasingly…
arXiv:2604.24661v5 Announce Type: replace Abstract: Visual control maps visual streams to actions in reinforcement learning and vision-language-action (VLA) policies…
arXiv:2610.40799v1 Announce Type: new Abstract: Across coding, web, and tool-use tasks, language-model agents are now routinely trained with the same recipe: supervised…
Associate Professor Cathy Wu uses reinforcement learning to help map out improvements to transportation and other multifaceted systems.
arXiv:2611.19190v1 Announce Type: new Abstract: Reinforcement learning (RL) can teach robot teams to navigate shared spaces, but trial and error causes collisions that…
arXiv:2611.05929v1 Announce Type: new Abstract: Off-policy deep reinforcement learning trains on random samples from a replay buffer of past experience. In physical…
arXiv:2611.13482v1 Announce Type: new Abstract: Predicting long action chunks, often spanning dozens of timesteps, has become a key ingredient in increasingly capable…
arXiv:2610.40902v1 Announce Type: new Abstract: Flapping-wing simulation must reconcile high-frequency wingbeat aerodynamics with the slower dynamics relevant to flight…
arXiv:2611.25197v1 Announce Type: new Abstract: Self-play reinforcement learning (RL) produces strong driving policies, but these policies are hard to verify or…
arXiv:2610.33508v1 Announce Type: new Abstract: Humanoid motion trackers reproduce diverse behaviors, but sharing a policy across bodies requires accounting for…
arXiv:2611.18003v1 Announce Type: new Abstract: Reinforcement learning has enabled robust humanoid locomotion, yet training recipes (rewards, architectures…
arXiv:2611.25251v1 Announce Type: new Abstract: Reinforcement learning (RL) can improve vision-language-action (VLA) policies beyond imitation, but humanoid exploration…
The work applies diffusion models to generate robot action sequences from observations.
Franka Research 3 is a 7-DoF torque-sensing robot arm aimed at research labs, offering 1 kHz low-level control and integration with ROS 2 and MATLAB.
The paper presents a mobile bimanual teleoperation system and imitation-learning experiments for household-style tasks.
The work studies closed-loop grasping learned from a large collection of robot interaction data.
The project aggregates robot demonstration data across institutions and studies models trained across different robot embodiments.
A set of methods that fit model behavior from data for tasks such as prediction, classification, perception or control.
arXiv:2610.33317v1 Announce Type: new Abstract: Deploying humanoid task policies requires reliable runtime monitors that anticipate unrecoverable falls in time to…
arXiv:2610.33227v1 Announce Type: new Abstract: The last decades have seen a revolution in autonomous robotics. Deep learning approaches and their hardware…
arXiv:2610.32277v1 Announce Type: new Abstract: Learned humanoid controllers are usually trained in one simulator and then checked in a second one before they go near…
arXiv:2611.13499v1 Announce Type: new Abstract: End-to-end navigation in cluttered environments relies on compact visual representations to support collision avoidance…
arXiv:2611.13307v1 Announce Type: new Abstract: Purpose: Robot-based haptic dyads, where two users exchange forces through robots, have been proposed for robot-assisted…
arXiv:2611.13685v1 Announce Type: new Abstract: In this paper, we investigate how human egocentric videos can benefit robot policy learning without action supervision…
arXiv:2611.19196v1 Announce Type: new Abstract: Imitation learning for visuomotor policies typically relies on teleoperated robot demonstrations, which are costly…
arXiv:2610.31845v1 Announce Type: new Abstract: Scaling policy learning requires broad experience and reliable supervision. Robot demonstrations provide high-quality…
arXiv:2611.05571v1 Announce Type: new Abstract: Robot imitation learning is limited by the scarcity of action-labeled data, while action-free demonstrations such as…
arXiv:2610.41336v1 Announce Type: new Abstract: Learning dexterous manipulation from human videos remains an open challenge, particularly in the context of transferring…
arXiv:2610.36374v1 Announce Type: new Abstract: Robotic agents operating in real-world environments must adapt continuously from a stream of multimodal demonstrations…
arXiv:2611.17658v1 Announce Type: new Abstract: Imitation policies for robot manipulation are often trained on noisy object positions, and they learn to discount the…
arXiv:2611.23502v1 Announce Type: new Abstract: We introduce TAG (Testbed for Autonomous Games), a compact, low-cost robotic platform for studying learning and control…
arXiv:2610.33486v1 Announce Type: new Abstract: A dexterous hand can change an object's pose without setting it down. This survey covers 113 learning-based methods that…
arXiv:2610.41320v1 Announce Type: new Abstract: Recent multimodal foundation models exhibit strong perception, reasoning, and in-context learning capabilities, but…
arXiv:2603.02511v2 Announce Type: replace Abstract: Sequential manipulation in clutter requires selecting which objects to remove and determining how to manipulate them…
arXiv:2611.23563v1 Announce Type: new Abstract: Many learning-based approaches for deformable linear objects (DLOs) rely on simplified deformation models that neglect…
arXiv:2611.23575v1 Announce Type: new Abstract: Low-cost tactile arrays have made touch a practical input for robot learning, but piezoresistive designs sense only…