Saved in:
Bibliographic Details
Main Authors: Strauch, Pascal, Müller, David, Christen, Sammy, Serifi, Agon, Grandia, Ruben, Knoop, Espen, Bächer, Moritz
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.10635
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914157190709248
author Strauch, Pascal
Müller, David
Christen, Sammy
Serifi, Agon
Grandia, Ruben
Knoop, Espen
Bächer, Moritz
author_facet Strauch, Pascal
Müller, David
Christen, Sammy
Serifi, Agon
Grandia, Ruben
Knoop, Espen
Bächer, Moritz
contents Despite recent advances in robust locomotion, bipedal robots operating in the real world remain at risk of falling. While most research focuses on preventing such events, we instead concentrate on the phenomenon of falling itself. Specifically, we aim to reduce physical damage to the robot while providing users with control over a robot's end pose. To this end, we propose a robot agnostic reward function that balances the achievement of a desired end pose with impact minimization and the protection of critical robot parts during reinforcement learning. To make the policy robust to a broad range of initial falling conditions and to enable the specification of an arbitrary and unseen end pose at inference time, we introduce a simulation-based sampling strategy of initial and end poses. Through simulated and real-world experiments, our work demonstrates that even bipedal robots can perform controlled, soft falls.
format Preprint
id arxiv_https___arxiv_org_abs_2511_10635
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Robot Crash Course: Learning Soft and Stylized Falling
Strauch, Pascal
Müller, David
Christen, Sammy
Serifi, Agon
Grandia, Ruben
Knoop, Espen
Bächer, Moritz
Robotics
Machine Learning
Despite recent advances in robust locomotion, bipedal robots operating in the real world remain at risk of falling. While most research focuses on preventing such events, we instead concentrate on the phenomenon of falling itself. Specifically, we aim to reduce physical damage to the robot while providing users with control over a robot's end pose. To this end, we propose a robot agnostic reward function that balances the achievement of a desired end pose with impact minimization and the protection of critical robot parts during reinforcement learning. To make the policy robust to a broad range of initial falling conditions and to enable the specification of an arbitrary and unseen end pose at inference time, we introduce a simulation-based sampling strategy of initial and end poses. Through simulated and real-world experiments, our work demonstrates that even bipedal robots can perform controlled, soft falls.
title Robot Crash Course: Learning Soft and Stylized Falling
topic Robotics
Machine Learning
url https://arxiv.org/abs/2511.10635