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:10.1007/978-3-031-21090-7_31

Cited by 21 publications

(3 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…RL has been used to imitate (22,23) or correct (32) motions obtained by solving TO problems. Conversely, model-based methods have been used to check the feasibility of learned high-level commands (27) or to track learned acceleration profiles (33). Compared with (32), we do not learn corrective joint torques around an existing WBC but, instead, learn the mapping from reference signals to joint positions in an end-to-end fashion.…”

Section: Introductionmentioning

confidence: 99%

DTC: Deep Tracking Control

Jenelten,

He,

Farshidian

et al. 2024

Sci. Robot.

View full text Add to dashboard Cite

Legged locomotion is a complex control problem that requires both accuracy and robustness to cope with real-world challenges. Legged systems have traditionally been controlled using trajectory optimization with inverse dynamics. Such hierarchical model-based methods are appealing because of intuitive cost function tuning, accurate planning, generalization, and, most importantly, the insightful understanding gained from more than one decade of extensive research. However, model mismatch and violation of assumptions are common sources of faulty operation. Simulation-based reinforcement learning, on the other hand, results in locomotion policies with unprecedented robustness and recovery skills. Yet, all learning algorithms struggle with sparse rewards emerging from environments where valid footholds are rare, such as gaps or stepping stones. In this work, we propose a hybrid control architecture that combines the advantages of both worlds to simultaneously achieve greater robustness, foot-placement accuracy, and terrain generalization. Our approach uses a model-based planner to roll out a reference motion during training. A deep neural network policy is trained in simulation, aiming to track the optimized footholds. We evaluated the accuracy of our locomotion pipeline on sparse terrains, where pure data-driven methods are prone to fail. Furthermore, we demonstrate superior robustness in the presence of slippery or deformable ground when compared with model-based counterparts. Last, we show that our proposed tracking controller generalizes across different trajectory optimization methods not seen during training. In conclusion, our work unites the predictive capabilities and optimality guarantees of online planning with the inherent robustness attributed to offline learning.

show abstract

Section: Introductionmentioning

confidence: 99%

DTC: Deep Tracking Control

Jenelten,

He,

Farshidian

et al. 2024

Sci. Robot.

View full text Add to dashboard Cite

show abstract

“…Mobile robots, and especially legged ones, have the capacity to evolve into multi-purpose machines, useful in many application scenarios, such as production sites, household services, remote inspection, and disaster search-andrescue [1], [2], [3]. Path planning is crucial in enabling legged robots to navigate autonomously and effectively complete the attributed tasks in various complex environments.…”

Section: Introductionmentioning

confidence: 99%

AAPP: An Accelerative and Adaptive Path Planner for Robots on GPU

Juan

et al. 2023

IEEE Trans. Comput.

View full text Add to dashboard Cite

In this work, we present DiPPeR, a novel and fast 2D path planning framework for quadrupedal locomotion, leveraging diffusion-driven techniques. Our contributions include a scalable dataset of map images and corresponding end-to-end trajectories, an image-conditioned diffusion planner for mobile robots, and a training/inference pipeline employing CNNs. We validate our approach in several mazes, as well as in real-world deployment scenarios on Boston Dynamic's Spot and Unitree's Go1 robots. DiPPeR performs on average 70 times faster for trajectory generation against both search based and data driven path planning algorithms with an average of 80% consistency in producing feasible paths of various length in maps of variable size, and obstacle structure.

show abstract

“…IHMC's Atlas robot [18] [17] also exhibits impressive line balancing capabilities on extremely thin beams using a whole body control method. There has also been attempts at using Reinforcement based methods have also been investigated by Xie et al [59] where a RL policy is trained to make a quadrupedal robot to perform a variety of maneuvers including balance beam locomotion.…”

Section: Narrow Path Walking : History and Motivationmentioning

confidence: 99%

Towards dynamic narrow path walking on NU's Husky

Venkatesh Krishnamurthy

View full text Add to dashboard Cite

This research focuses on enabling Northeastern University's Husky, a multi-modal quadrupedal robot, to navigate narrow paths akin to various animals in nature. The Husky is equipped with thrusters to stabilize its body during dynamic maneuvers, addressing challenges inherent in aerial-legged systems. The approach involves modeling the robot as HROM (Husky Reduced Model) and creating an optimal control framework using linearized dynamics for narrow path walking. The thesis introduces a gait scheduling method to generate an open-loop walking gait and validates these gaits through a high-fidelity Simscape simulation. Experimental results of the open-loop walking are presented, accompanied by potential directions for advancing this robotic system.

show abstract

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

Cited by 21 publications

References 42 publications

DTC: Deep Tracking Control

DTC: Deep Tracking Control

AAPP: An Accelerative and Adaptive Path Planner for Robots on GPU

Towards dynamic narrow path walking on NU's Husky

Contact Info

Product

Resources

About