aiqbench analysis
This study demonstrates a unified neural network controller for a planetary rover that eliminates the need for explicit terrain classification by consolidating policies across heterogeneous environments. The approach achieves zero-shot transfer to physical hardware through domain randomization, resulting in a 37% reduction in cost of transport on dry sand and enabling mobility on wet sand where passive systems fail. For robotics practitioners, the key takeaway is the successful integration of high-fidelity terramechanics simulations with reinforcement learning to create robust, adaptive locomotion strategies without manual controller switching.
Learning All-Terrain Locomotion for a Planetary Rover with Actively Articulated Suspension
arXiv:2606.06790v1 Announce Type: cross Abstract: This paper presents ERNEST, a four-wheeled planetary rover concept equipped with a two-degree-of-freedom Active Gimbal Suspension that combines yaw and roll actuation to enable wheel reconfiguration, steering, and active load redistribution. A single neural network controller, trained to track a desired path across challenging terrain, fully unlocks the capabilities of this actuated suspension system for autonomous obstacle negotiation. A reinforcement learning framework is developed using the high-fidelity DARTS simulation engine, which combines rigid-contact dynamics and Bekker-Wong terramechanics, enabling the emergence of locomotion strategies adapted to loose-soil conditions. To obtain a single unified controller across heterogeneous terrains, a policy consolidation strategy merges the experience of terrain-specialized agents into one neural network, eliminating the need for explicit terrain classification and controller switching. The resulting controller operates on a combination of proprioceptive and exteroceptive feedback, including sparse stereo-derived terrain elevation, chassis attitude, joint states, and force-torque measurements. Zero-shot transfer to the physical rover is achieved through domain randomization, sensor noise injection, and model-to-real system identification. Experimental results demonstrate autonomous traversal of rock fields, a bump trap, a wheel-high step, sand ripples, and sandy slopes. On a 20{\deg} sandy slope, the learned controller reduces the cost of transport by 37% on dry sand despite the additional actuation, and achieves superior performance on wet sand where the passive suspension becomes completely immobilized.
Abstract:This paper presents ERNEST, a four-wheeled planetary rover concept equipped with a two-degree-of-freedom Active Gimbal Suspension that combines yaw and roll actuation to enable wheel reconfiguration, steering, and active load redistribution. A single neural network controller, trained to track a desired path across challenging terrain, fully unlocks the capabilities of this actuated suspension system for autonomous obstacle negotiation. A reinforcement learning framework is developed using the high-fidelity DARTS simulation engine, which combines rigid-contact dynamics and Bekker-Wong terramechanics, enabling the emergence of locomotion strategies adapted to loose-soil conditions. To obtain a single unified controller across heterogeneous terrains, a policy consolidation strategy merges the experience of terrain-specialized agents into one neural network, eliminating the need for explicit terrain classification and controller switching. The resulting controller operates on a combination of proprioceptive and exteroceptive feedback, including sparse stereo-derived terrain elevation, chassis attitude, joint states, and force-torque measurements. Zero-shot transfer to the physical rover is achieved through domain randomization, sensor noise injection, and model-to-real system identification. Experimental results demonstrate autonomous traversal of rock fields, a bump trap, a wheel-high step, sand ripples, and sandy slopes. On a 20° sandy slope, the learned controller reduces the cost of transport by 37% on dry sand despite the additional actuation, and achieves superior performance on wet sand where the passive suspension becomes completely immobilized.
Robotics (cs.RO); Machine Learning (cs.LG); Systems and Control (eess.SY)
arXiv:2606.06790 [cs.RO]
(or arXiv:2606.06790v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2606.06790 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 317 words
Related coverage
TAGA: Terrain-aware Active Gaze Learning for Generalizable Agile Humanoid Locomotion
arXiv · Sep 28, 2026
TAGA: Terrain-aware Active Gaze Learning for Generalizable Agile Humanoid Locomotion
arXiv · Jun 5, 2026
Learning Air-Ground Motion Control with Temporal Mode Switching and Cross-Terrain Tracking
arXiv · Sep 24, 2026
Physical Imitation Learning: Distilling Control Policies into Passive Elasticity
arXiv · Jun 17, 2026