KIMM develops elephant-trunk-inspired robot hand that can grasp objects from needles to boxes

The Korea Institute of Machinery and Materials (KIMM) announced on October 20, 2022, that it had developed a gripper robot capable of grasping objects ranging from extremely thin items such as needles and pins to boxes. The team was led by Song Seong-hyeok, a senior researcher in the Robotics and Mechatronics Division of the AI Robot Research Division. The robot is designed to combine pinching at its tip, like an elephant’s trunk, with suction.

KIMM’s announcement points to an effort to handle both fragile small objects and larger items such as boxes with the same end-effector. In factory and logistics operations, stable gripping is a prerequisite for the entire process, so this type of design can directly affect whether a robot can be deployed.

What is new in this announcement is the integration of pinching and suction in one mechanism to expand the range of objects it can handle to include both thin items and boxes. As an improvement to a robot hand, this extends an established line of development, but the design combines a flexible structure inspired by an elephant’s trunk with vacuum channels, distinguishing it from a simple claw-style gripper. The important point is not the showiness of its movements but the mechanism’s design, which aims to move objects stably without damaging them. Related articles from Physical AI News, KIMM opens development site for humanoid “KAIROS” and KIMM accelerates robot development with robot-task AI system, show that KIMM is opening its robot-development sites and building out learning systems. However, the source for this announcement is limited to the gripping mechanism and does not describe a direct connection to humanoids or task AI. It is therefore reasonable to view this as an expansion of the gripping mechanism as part of KIMM’s broader robot development work, while avoiding the assumption that results from separate projects form a single effort. From an industry perspective, this technology affects the end-effector that physically grasps and moves objects, rather than perception. In other words, it improves the component that accommodates differences in the shapes of objects and helps reduce failures at the start of a process. The source, however, gives no figures for durability, gripping speed, control method, or the weights and materials it can handle, so its effectiveness in actual operations remains uncertain. The next points to check are the test conditions, the limits of the objects it can grasp, and how it switches between suction and pinching.

The source does not identify any suppliers or customers in Japan, so the direct impact on Japan cannot be determined.