ETD

• grasping
• multi-object

The concept of grasping is natural for humans, who can manipulate objects while immediately perceiving properties such as shape, material, weight, and surface characteristics. For robots, however, grasping is a more complex process that requires collecting and processing information to perform effective and reliable manipulation.
A fundamental concept is grip, defined as the ability of a robot to grasp, hold, and manipulate an object in a controlled manner. Grip depends on both the hardware, such as the gripper or robotic hand, and the control system, which manages force application, positioning, and manipulation stability.
Traditional robotic grasping mainly focused on single-object grasping, where objects are picked one at a time. However, researchers extended this idea by asking whether robots could manipulate multiple objects simultaneously. This led to the development of Multi-Object Grasping (MOG), a technique in which several objects are grasped and transferred within a single action.
In MOG applications, the robot must determine which objects to pick together, regulate the applied force, and maintain stable manipulation while avoiding collisions. The objective is to improve productivity and efficiency by reducing the number of manipulation cycles, especially in industrial and cluttered environments where objects are rarely isolated.