Physical AI News (Japan)

Physical AI, robotics and compute — reported from Japan.

日本語
· Research & technology· 日本語で読む

UC3M and UPNA Test New Soft Robotic Gripper That Can Reposition Objects While Holding Them

Researchers at Spain’s UC3M and UPNA developed a soft robotic gripper that can shift or reposition an object between its fingers after grasping it. The research was reportedly published in the scientific journal Bioinspiration & Biomimetics. Unlike rigid industrial robots, the gripper is designed to handle fragile or irregularly shaped objects, according to reports.

What happened

Researchers at Spain’s Universidad Carlos III de Madrid (UC3M) and Universidad Pública de Navarra (UPNA) developed a soft robotic gripper that can not only grasp and hold an object but also shift or reposition it between its fingers without losing its grip, techxplore.com reported on October 4, 2026. The research was reportedly published in the scientific journal Bioinspiration & Biomimetics.

Why it matters

Researchers say a soft gripper that can move an object while holding it could expand the range of tasks possible in processes involving fragile or irregularly shaped objects—tasks that conventional rigid gripping mechanisms have difficulty handling. Its intended use in operations near people also sets it apart from a simple grasping demonstration.

Analysis

The focus here is not simply a device that uses soft materials to grasp objects, but one that can move an object between its fingers after grasping it. Soft grippers are considered well suited to fragile objects and objects with varied shapes, but they have been seen as more limited than rigid industrial robots when it comes to repositioning objects with their fingertips and controlling their orientation. In that sense, the report advances a research theme: how to combine “softness” with dexterity. It is not merely a materials improvement. Publicly available information does not indicate the price or prospects for mass production, so for now the work appears closer to performance testing of a research prototype. We have no related articles, but in context, the next measure of progress for soft robotics is whether it can move beyond “not breaking objects” to manipulation after grasping. Research on flexible fingers often stops at picking or safe contact, but this report is significant because it describes a step beyond that: changing an object’s position while continuing to hold it. Public information does not confirm what kinds of objects can be repositioned reliably or at what speeds and loads. Those details would help clarify the potential applications beyond laboratory demonstrations. Views also differ on how this technology relates to industrial robots. Rigid robots are widely used in automotive and electronics processes because of their high precision and stiffness, while soft gripping mechanisms of this kind are more likely to complement them in work involving difficult-to-handle objects or close interaction with people than to replace them. Accordingly, the main area of competition is likely to shift from general-purpose gripping performance to the range of objects a system can adapt to, ease of control, durability and ease of safety certification. Points to watch include: (1) which objects can be repositioned, (2) the achievable load, speed and cycle life, and (3) the control systems and sensor configurations needed to move from a research prototype to implementation.

Implications for Japan

Public information does not indicate a direct connection to Japanese companies or manufacturing sites. However, the evaluation criteria for soft gripping mechanisms of this kind could also affect the design of collaborative robots and peripheral equipment in Japan, particularly for light-duty tasks involving wide variations in object shape or processes where close interaction with people is difficult to avoid.

Sources

Translated from the Japanese original: /news/uc3m-to-upna-buttai-wotsukandamama-ugoka-seru-nantai-robotto-42085b. The Japanese version is authoritative if they differ.