Adaptive legged robots adjust to unknown objects and terrain in real time
Adaptive-MHE : A Sampling-Based Adaptive MPC for Legged Loco-Manipulation via Moving Horizon Estimation
Robotics
Summary
Robots with legs can walk on different surfaces, but handling objects while walking is difficult because they don’t know details about the objects or ground. The authors created a method called Adaptive-MHE that helps robots learn these details as they move, by matching real movements with simulated ones repeatedly. This approach lets robots adjust their actions better in changing environments, improving their walking and object handling. Tests showed that this method works better than previous techniques and almost as well as if the robot knew all the details beforehand.
What this means in practice
- •For robotics engineers: Develop legged robots that adapt in real time to unknown object and terrain properties to improve walking and manipulation abilities.
- •For automation system designers: Create adaptive control systems that maintain performance despite changes in physical environments or payloads for better operational reliability.
Authors
Hossein Keshavarz, Alejandro Ramirez-Serrano, Majid Khadiv
Abstract
Legged robots have demonstrated a remarkable ability to traverse various terrains, yet generating effective loco-manipulation behaviors remains challenging. A key difficulty is that object and terrain parameters are typically unknown to the robot, and mismatches between these parameters and their simulated counterparts introduce a sim-to-real gap that degrades control performance. Classical system identification (Sys-ID) methods often assume differentiable dynamics, an assumption that does not hold for contact-rich legged systems. Sampling-based Sys-ID avoids this restriction by directly matching simulated and recorded state trajectories through massively parallel rollouts, but existing approaches are typically applied offline and do not adapt as environmental conditions change. We present Adaptive-MHE an online sampling-based Sys-ID framework, based on moving horizon estimation (MHE), that estimates the physical parameters of objects and terrain in the environment (e.g., mass, friction) and couples this estimate with a sampling-based model predictive controller, enabling adaptive loco-manipulation in changing and uncertain environments. In simulation and hardware experiments, our framework consistently outperforms baselines and matches the performance of a controller with access to ground-truth parameters.