Stage-focused force sensing improves robot contact tasks
DEFT-VLA: Stage-Localized Force Specialization for Contact-Rich Robotic Manipulation
Robotics
Summary
Robots often struggle with tasks that involve touching or manipulating objects because different senses are more useful at different stages. For example, vision helps to locate and move toward an object, while feeling the force is important once contact is made. The authors developed a method called DEFT-VLA that splits the robot’s skills into two parts: one for using vision before touching, and another that focuses on force sensing during contact. This approach helps robots perform better at tasks that need precise physical interaction, like attaching pipette tips or wiping surfaces.
What this means in practice
- •For robotics engineers: Develop robotic systems that handle delicate tasks better by adapting force sensing only during physical contacts, improving task success rates.
- •For automation system integrators: Integrate stage-specific sensor processing in industrial robots to enhance precision in assembly tasks involving contact without compromising vision-based positioning.
Authors
Zan Huang, Yibo Qiu, Kan Ding, Xuanping Zhang, Ao Tian, Ronald X Xu, Mingzhai Sun
Abstract
Contact-rich manipulation poses a stage-dependent sensory conflict: vision and proprioception are essential for localization and approach, when force is largely uninformative and potentially distracting, whereas force becomes critical after contact, when irrelevant visual details and correlated robot states may become shortcuts. We propose DEFT-VLA (Dual-Expert Force-specialized Training for Vision-Language-Action Manipulation), which retains the task-adapted base expert for vision-dominated non-contact behavior and introduces a force-aware contact expert for critical contact stages, localizing force-conditioned adaptation to where it is needed. We further introduce a contact-localized force-attending curriculum that applies force-attending post-training only to the contact expert, progressively restoring vision and masking state tokens to suppress shortcuts. Real-robot experiments on pipette tip attachment, bimanual pipette aspiration, and whiteboard wiping show that DEFT-VLA effectively exploits force feedback to improve contact-stage performance while maintaining strong non-contact behavior and robustness to held-out grasp heights. It improves overall task success by 20-40 percentage points over the strongest baseline on each task.