The human hand is capable of handling an object through the use of a combination of senses which determines the texture of the object. This enables adjustment of one’s grip by either tightening or releasing the object. On the other hand, a robot requires a gripper which is attached to a manipulator to be able to perform this function.
However, when a robot gripper is used to hold an object, it becomes difficult to accurately decide how to handle the object. This is because a robot gripper does not have the senses to accurately decide what kind of hold to apply to what object. The importance of a robot gripper cannot be underestimated. It is the physical part that connects the robot arm to the object. It is responsible for picking and placing an object. The safety of an object, therefore, depends on the type of gripper used. There are various types of grippers that are appropriate for the various types of items to be carried. This function makes it important to have a robot gripper which can adjust its grip to any kind of material.
To do this, it would require the senses to determine when an object requires a delicate touch. A sensor system to achieve this has been developed by ETH researchers. The model includes an upward-pointing camera, microbeads that are colored and skin made from silicone.
The changing order of the microbeads is picked up by the camera when contact is made between an object and the skin. Experiments were carried out using different sizes of objects which were placed on different points of the sensor skin with varying amounts of force applied. The changing order of the microbeads with the different sizes and amounts of applied force onto the sensor was then recorded with a machine. This was done using Artificial intelligence to enable the sensor system to precisely identify the application of lateral and vertical forces.
According to Carlo Sferrazza a Ph.D. student, the traditional sensor arrangement only recognizes pressure applied to a particular point. However, the robotic skin can only tell apart the difference between the various kind of pressure applied to the surface of the sensor and the trajectory of the applied pressure. This it achieves with a high degree of accuracy and by producing sharp and finely defined images.
Currently, the measurement for the model is at 5 by 5 cm or 1.96 by 1.96 inches and is 1. 7cm (0.66 inches) thick. Efforts are being made by researchers to actualize a sensor system that will cover a broad surface, encompass a lot of cameras, and not as thick as the current model.
The research team views the potential in providing robot grippers with the ability to handle objects by adjusting their grips according to the type of object. When we pick up an object, we can adjust our grip if that object proves to be a slippery customer The team also believes that this sensor system would be relatively cheap to reproduce. And also, that it would be useful to athletes and fitness fanatics and hopefully in Tech.
