VacuumVLA: Two skills, one robot hand
Robots guided by Vision-Language-Action (VLA) models are getting good at everyday tasks — but most still grab with simple two-finger claws. That limits them on flat, slippery, or handle-less surfaces.
VacuumVLA adds a low-cost twist: a single end-effector that combines a standard gripper with a vacuum suction tool. It can switch modes on the fly or use both together, giving robots two ways to hold and move things.
- Pick up flat items and wipe glass or tables
- Open drawers without handles
- Stabilize objects while cleaning or manipulating
Plugged into state-of-the-art VLA systems (DexVLA and Pi0), this hybrid hand unlocked tasks that regular grippers couldn’t handle—without exotic hardware.
Two skills, one hand. More reach, more tasks.
Designs and control code will be released. Read the paper: https://arxiv.org/abs/2511.21557v1
Paper: https://arxiv.org/abs/2511.21557v1
Register: https://www.AiFeta.com
Robotics AI VLA Manipulation Hardware OpenSource Research Robots
