aiqbench analysis
This research addresses a critical bottleneck in deploying Vision-Language-Action (VLA) models by introducing a self-compensating mechanism that adapts to real-time robot execution errors caused by mechanical wear or payload changes. By utilizing an online update process based on action residuals rather than external rewards, the method significantly boosts task success rates across diverse physical conditions and unseen objects. The introduction of RoboStress provides a valuable simulation benchmark for rigorously testing VLA robustness against complex factors like friction and backlash before physical deployment.
Taming VLAs under Robot Execution Errors: Self-Compensation and Stress Testing
arXiv:2609.37334v1 Announce Type: new Abstract: Vision-language-action (VLA) policies often fail when a robot's executed motion deviates from their commanded action. Such execution errors arise from the robot's mechanics and operating conditions, such as wear and payload changes. We propose self-compensating VLA, a deployment-time adaptation method that enables a VLA policy to pre-compensate for the robot's execution errors when generating commands. Without task rewards or labels, it updates the policy online using the residual between the action commanded by a VLA and the motion executed by the robot. To stress-test VLA robustness across execution conditions that are impractical to cover with physical robots alone, we introduce RoboStress, a controlled simulation benchmark. It combines established joint-level models of friction, backlash, compliance, and gravity-compensation error into seven deployment scenarios whose execution errors depend on the robot's state and motion history. On RoboStress, self-compensating VLA achieves higher average task success than both the base policies and methods that build in robustness during training. On two physical robot arms with different usage histories, it raises the average task success rate by more than 30 percentage points on each arm, and the gains extend to objects not seen in the task demonstrations.
Abstract:Vision-language-action (VLA) policies often fail when a robot's executed motion deviates from their commanded action. Such execution errors arise from the robot's mechanics and operating conditions, such as wear and payload changes. We propose self-compensating VLA, a deployment-time adaptation method that enables a VLA policy to pre-compensate for the robot's execution errors when generating commands. Without task rewards or labels, it updates the policy online using the residual between the action commanded by a VLA and the motion executed by the robot. To stress-test VLA robustness across execution conditions that are impractical to cover with physical robots alone, we introduce RoboStress, a controlled simulation benchmark. It combines established joint-level models of friction, backlash, compliance, and gravity-compensation error into seven deployment scenarios whose execution errors depend on the robot's state and motion history. On RoboStress, self-compensating VLA achieves higher average task success than both the base policies and methods that build in robustness during training. On two physical robot arms with different usage histories, it raises the average task success rate by more than 30 percentage points on each arm, and the gains extend to objects not seen in the task demonstrations.
Robotics (cs.RO)
arXiv:2609.37334 [cs.RO]
(or arXiv:2609.37334v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.37334 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 · 284 words
Related coverage
CycleVLA: Proactive Self-Correcting Vision-Language-Action Models via Subtask Backtracking and Minimum Bayes Risk Decoding
arXiv · Jul 29, 2026
The Machines that Make the Machines
NVIDIA · Oct 7, 2026
News
Quantum Computing Report · Jun 7, 2026
Vision-Based Early Fault Diagnosis and Self-Recovery for Strawberry Harvesting Robots
arXiv · Jun 9, 2026