GLiDE: Generalizable Quadrupedal Locomotion in Diverse Environments with a Centroidal Model

Xie, Zhiqiang; Da, Xingye; Babich, B. N.; Garg, Animesh; Panne, Michiel van de

doi:10.48550/arxiv.2104.09771

Cited by 13 publications

(13 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, these low-dimensional models depend on a human-designed reference motion, which limits their application to a few gaits which are robotic and not naturallooking. Other works combine model-based controllers with reinforcement learning [24], [25] to generate desired CoM acceleration, or use learned dynamics models [26] for planning. In this work, we adapt the model-based controller from [2] and use animal motion trajectories as the reference motion to generate diverse, agile, natural motions on an A1 quadrupedal robot.…”

Section: B Model-based Legged Locomotion Controlmentioning

confidence: 99%

“…In this work, we adapt the model-based controller from [2] and use animal motion trajectories as the reference motion to generate diverse, agile, natural motions on an A1 quadrupedal robot. Instead of using RL like [24], we use trajectory optimization to improve the performance of the model-based controller in simulation, and transfer optimized reference trajectory to the real robot.…”

Section: B Model-based Legged Locomotion Controlmentioning

confidence: 99%

See 1 more Smart Citation

FastMimic: Model-based Motion Imitation for Agile, Diverse and Generalizable Quadrupedal Locomotion

Li¹,

Won²,

Ha³

et al. 2021

Preprint

View full text Add to dashboard Cite

Robots operating in human environments need a variety of skills, like slow and fast walking, turning, and sidestepping. However, building robot controllers that can exhibit such a large range of behaviors is challenging, and unsolved. We present an approach that uses a model-based controller for imitating different animal gaits without requiring any realworld fine-tuning. Unlike previous works that learn one policy per motion, we present a unified controller which is capable of generating four different animal gaits on the A1 robot. Our framework includes a trajectory optimization procedure that improves the quality of real-world imitation. We demonstrate our results in simulation and on a real 12-DoF A1 quadruped robot. Our result shows that our approach can mimic four animal motions, and outperform baselines learned per motion.

show abstract

Section: B Model-based Legged Locomotion Controlmentioning

confidence: 99%

Section: B Model-based Legged Locomotion Controlmentioning

confidence: 99%

FastMimic: Model-based Motion Imitation for Agile, Diverse and Generalizable Quadrupedal Locomotion

Li¹,

Won²,

Ha³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Unlike our approach, RLOC still uses a WBC which may limit the types of movements that can be selected by the perceptive foothold setting policy. GLiDE [18] learns an RL policy for the centroidal dynamics directly. This system was shown to traverse narrow beams and small stepping stones.…”

Section: B Combining Rl and Planning For Locomotionmentioning

confidence: 99%

Learning Coordinated Terrain-Adaptive Locomotion by Imitating a Centroidal Dynamics Planner

Brakel¹,

Bohez²,

Hasenclever³

et al. 2021

Preprint

View full text Add to dashboard Cite

Dynamic quadruped locomotion over challenging terrains with precise foot placements is a hard problem for both optimal control methods and Reinforcement Learning (RL). Non-linear solvers can produce coordinated constraint satisfying motions, but often take too long to converge for online application. RL methods can learn dynamic reactive controllers but require carefully tuned shaping rewards to produce good gaits and can have trouble discovering precise coordinated movements. Imitation learning circumvents this problem and has been used with motion capture data to extract quadruped gaits for flat terrains. However, it would be costly to acquire motion capture data for a very large variety of terrains with height differences. In this work, we combine the advantages of trajectory optimization and learning methods and show that terrain adaptive controllers can be obtained by training policies to imitate trajectories that have been planned over procedural terrains by a non-linear solver. We show that the learned policies transfer to unseen terrains and can be fine-tuned to dynamically traverse challenging terrains that require precise foot placements and are very hard to solve with standard RL.

show abstract

“…Alternatively, researchers [4], [19] have investigated learning policies directly from realworld experience, which can intrinsically overcome sim-toreal gaps. Sample efficiency is a critical challenge for deep RL approaches, which can be improved by leveraging modelbased control strategies [17], [20], [21]. Recently, Lee et al [22] employed a teacher-student framework for training their agent.…”

Section: Related Workmentioning

confidence: 99%

Solving Challenging Control Problems Using Two-Staged Deep Reinforcement Learning

Sontakke

2021

Preprint

View full text Add to dashboard Cite

We present a two-staged deep reinforcement learning algorithm for solving challenging control problems. Deep reinforcement learning (deep RL) has been an effective tool for solving many high-dimensional continuous control problems, but it cannot effectively solve challenging problems with certain properties, such as sparse reward functions or sensitive dynamics. In this work, we propose an approach that decomposes the given problem into two stages: motion planning and motion imitation. The motion planning stage seeks to compute a feasible motion plan with approximated dynamics by directly sampling the state space rather than exploring random control signals.Once the motion plan is obtained, the motion imitation stage learns a control policy that can imitate the given motion plan with realistic sensors and actuations. We demonstrate that our approach can solve challenging control problemsrocket navigation and quadrupedal locomotion -which cannot be solved by the standard MDP formulation. The supplemental video can be found at: https://youtu.be/FYLo1Ov_8-g

show abstract

GLiDE: Generalizable Quadrupedal Locomotion in Diverse Environments with a Centroidal Model

Cited by 13 publications

References 44 publications

FastMimic: Model-based Motion Imitation for Agile, Diverse and Generalizable Quadrupedal Locomotion

FastMimic: Model-based Motion Imitation for Agile, Diverse and Generalizable Quadrupedal Locomotion

Learning Coordinated Terrain-Adaptive Locomotion by Imitating a Centroidal Dynamics Planner

Solving Challenging Control Problems Using Two-Staged Deep Reinforcement Learning

Contact Info

Product

Resources

About