Hybrid Roller-Jamming Gripper for Object Acquisition and Retention Under Pose Uncertainty
This paper presents a hybrid roller-jamming gripper that combines active roller-driven intake with vacuum-induced granular jamming to achieve robust object acquisition and retention under significant pose uncertainty, demonstrating a 96.7% success rate across 840 grasp trials compared to significantly lower performance in ablation studies using either mechanism alone.
Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Imagine a robot trying to pick up a pill bottle from a cluttered nightstand or a sock from a messy floor. In a perfect laboratory, the robot knows exactly where the object is and how it is lying. But in a real home, objects are often hidden, tilted, or partially blocked from view. When a robot reaches out based on a guess, it frequently misses the center of the object, making contact with only a corner or a side. This imperfect first touch is a major hurdle for domestic robots. To solve this, engineers have developed two main types of robotic hands. One type uses soft, flexible fingers that wrap around objects, while another uses a special technique called granular jamming, where a bag of tiny particles is sucked tight with a vacuum to become rigid and hold on. However, the soft hands can struggle to pull a misaligned object into a secure grip, and the jamming hands need a good, full contact before they can stiffen up effectively. If the initial touch is poor, the jamming hand might just lock onto a tiny, unstable point and drop the item.
Researchers at Waseda University have built a new robotic gripper that combines the best of both worlds into a single device. This hybrid gripper features two fingers, each ending in a large, spherical roller made of a soft, air-filled membrane filled with tiny particles. The device operates in two distinct phases to handle the uncertainty of a messy environment. First, when the fingers touch an object, even if the contact is off-center or incomplete, the rollers begin to spin inward. This rolling motion actively drags the object toward the center of the gripper, much like a conveyor belt pulling a package into place. Once the object is drawn in and the contact is secure, the second phase begins: a vacuum is applied to the rollers. This sucks the air out of the particle-filled membranes, causing them to instantly stiffen and lock the object firmly in place. The result is a hand that can first fix a bad approach and then hold on tight.
To test if this idea works, the team mounted their prototype on a seven-jointed robotic arm and ran a massive series of trials. They selected eight very different objects to represent the variety found in a home: a pen, a marble, a large gallon jug, a plastic cup, a sheet of paper, a T-shirt, a pressure gauge, and a cable. For each object, they deliberately introduced errors. In some tests, they shifted the starting position of the object by up to twenty percent of its width, simulating a misjudged approach. In others, they tilted the robot's hand by ten degrees in various directions to mimic a clumsy angle of approach. They repeated these tests hundreds of times, totaling 840 separate attempts. The results were striking. The hybrid gripper succeeded in 812 of those 840 trials. It was particularly robust against position errors, succeeding in 215 out of 216 attempts where the object was shifted. It also handled the tilted approaches well, succeeding in 597 out of 624 trials.
To understand exactly how much each part of the system contributed, the researchers ran a second set of experiments where they turned off specific features. When they disabled the rolling motion and let the gripper rely only on the jamming, the success rate dropped significantly, especially for thin or irregular objects like the pen and the cable. Without the rollers to pull the object into a better position, the jamming mechanism often locked onto a weak, partial contact and failed. Conversely, when they disabled the jamming and let the gripper rely only on the rolling, it could often grab the object but struggled to hold it securely, particularly with heavy items like the gallon jug. The rolling motion was excellent at acquisition, but the jamming was essential for retention. The full hybrid system, which uses rolling to fix the grip and jamming to lock it, proved to be the most reliable approach.
The study suggests that combining active movement with passive stiffening is a powerful strategy for robots operating in unpredictable environments. While the current prototype is still a laboratory device that requires careful calibration for each specific object, the core mechanism has been validated. The researchers found that the ability to actively draw an object inward before stiffening the grip allows the robot to recover from imperfect starts that would cause other grippers to fail. This work provides a concrete step toward robots that can navigate the messy, unstructured reality of a human home, turning a clumsy reach into a secure hold.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.