Advances, challenges, and opportunities for legged robots

2026/07/31 by Jonas Frey, Matías Mattamala, Hae-Won Park +5
Computer Science · #cs.RO

paper · pdf · doi:10.1126/scirobotics.aee0787

published as Science Robotics, Vol. 11, Issue 116 (2026) · Accepted for publication in Science Robotics. This is the author's version of the work. It is posted here by permission of the AAAS for personal use, not for redistribution. The definitive version was published in Science Robotics on July 29, 2026, DOI: 10.1126/scirobotics.aee0787

arxiv created 2026/07/31 · arxiv updated 2026/08/03

Abstract

Humanoid and quadrupedal robots have the potential to revolutionize the way we work, interact, and coexist with intelligent machines. To understand their effects on society and how they can enable scientific discovery, we assess the current capabilities of these systems along hardware, locomotion, autonomy, data, and applications. We identify recent advances and key open challenges that must be overcome to enable widespread adoption and new use cases for legged robots. Last, we provide an outlook on the future of legged robots, exploring their ethical considerations, economic potential, policy implications, and broader societal effects.

Citations

• SAM 3D: 3Dfy Anything in Images
• GaussGym: An open-source real-to-sim framework for learning locomotion from pixels
• Track Any Motions under Any Disturbances
• BeyondMimic: From Motion Tracking to Versatile Humanoid Control via Guided Diffusion
• Enter the Mind Palace: Reasoning and Planning for Long-term Active Embodied Question Answering
• A Careful Examination of Large Behavior Models for Multitask Dexterous Manipulation
• KungfuBot: Physics-Based Humanoid Whole-Body Control for Learning Highly-Dynamic Skills
• Attention-Based Map Encoding for Learning Generalized Legged Locomotion
• Design of a 3-DOF Hopping Robot with an Optimized Gearbox: An Intermediate Platform Toward Bipedal Robots
• TWIST: Teleoperated Whole-Body Imitation System
• Learned Perceptive Forward Dynamics Model for Safe and Platform-aware Robotic Navigation
• Boxi: Design Decisions in the Context of Algorithmic Performance for Robotics
• High-Performance Reinforcement Learning on Spot: Optimizing Simulation Parameters with Distributional Measures
• Holistic Fusion: Task- and Setup-Agnostic Robot Localization and State Estimation With Factor Graphs
• GR00T N1: An Open Foundation Model for Generalist Humanoid Robots
• VGGT: Visual Geometry Grounded Transformer
• Learning Humanoid Standing-up Control across Diverse Postures
• ASAP: Aligning Simulation and Real-World Physics for Learning Agile Humanoid Whole-Body Skills
• NaVILA: Legged Robot Vision-Language-Action Model for Navigation
• HOVER: Versatile Neural Whole-Body Controller for Humanoid Robots
• One Policy to Run Them All: an End-to-end Learning Approach to Multi-Embodiment Locomotion
• PIE: Parkour with Implicit-Explicit Learning Framework for Legged Robots
• Co-RaL: Complementary Radar-Leg Odometry with 4-DoF Optimization and Rolling Contact
• Humanoid Parkour Learning
• OpenVLA: An Open-Source Vision-Language-Action Model
• DiffuseLoco: Real-Time Legged Locomotion Control with Diffusion from Offline Datasets
• AutoInspect: Towards Long-Term Autonomous Industrial Inspection
• CaT: Constraints as Terminations for Legged Locomotion Reinforcement Learning
• Resilient Legged Local Navigation: Learning to Traverse with Compromised Perception End-to-End
• Robot Parkour Learning
• 3D Gaussian Splatting for Real-Time Radiance Field Rendering
• 3D Gaussian Splatting for Real-Time Radiance Field Rendering
• ANYmal Parkour: Learning Agile Navigation for Quadrupedal Robots
• Optimizing Bipedal Locomotion for The 100m Dash With Comparison to Human Running
• Fast Traversability Estimation for Wild Visual Navigation
• Goal-Conditioned Imitation Learning using Score-based Diffusion Policies
• Diffusion Policy: Visuomotor Policy Learning via Action Diffusion
• Diffusion policy: Visuomotor policy learning via action diffusion
• Real-World Humanoid Locomotion with Reinforcement Learning
• Real-world humanoid locomotion with reinforcement learning
• DayDreamer: World Models for Physical Robot Learning
• Learning Agile Skills via Adversarial Imitation of Rough Partial Demonstrations
• Sim-to-Lab-to-Real: Safe reinforcement learning with shielding and generalization guarantees
• Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement\n Learning
• Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learning
• Safe Learning in Robotics: From Learning-Based Control to Safe Reinforcement Learning
• RMA: Rapid Motor Adaptation for Legged Robots
• STEP: Stochastic Traversability Evaluation and Planning for Risk-Aware Off-road Navigation
• Learning Transferable Visual Models From Natural Language Supervision
• Dynamic Walking: Toward Agile and Efficient Bipedal Robots
• Learning by Cheating
• Robots and Jobs: Evidence from US Labor Markets
• A Review of Robot Learning for Manipulation: Challenges, Representations, and Algorithms
• AMASS: Archive of Motion Capture as Surface Shapes
• Learning agile and dynamic motor skills for legged robots
• Paired Open-Ended Trailblazer (POET): Endlessly Generating Increasingly Complex and Diverse Learning Environments and Their Solutions
• Learning to Walk via Deep Reinforcement Learning
• Mode-adaptive neural networks for quadruped motion control
• Sim-to-Real: Learning Agile Locomotion For Quadruped Robots
• Gaze and the Control of Foot Placement When Walking in Natural Terrain
• DeepMimic: Example-Guided Deep Reinforcement Learning of Physics-Based Character Skills
• Maximum a Posteriori Policy Optimisation
• Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
• Proximal Policy Optimization Algorithms
• Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World
• Robots in aged care: a dystopian future?
• Atlas
• Isaac Lab: A GPU-Accelerated Simulation Framework for Multi-Modal Robot Learning
• Gemini Robotics: Bringing AI into the Physical World
• Dual-arm Motion Generation for Repositioning Care based on Deep Predictive Learning with Somatosensory Attention Mechanism