In this paper, an object classification system by thermal conductivity is introduced. This system comprises a module of a flexible thermoelectric device (TED) and a resistance temperature detector sensor (RTD). The module is integrated with a silicone finger cot and is equipped onto a robot hand. Before grasping an object, the TED on the robot hand is heated to a specific temperature in degrees Celsius higher than room temperature. While the robot hand grasps an object, heat generated by the TED is transferred to the object. Sequentially, the RTD sensor detects heat variation of the object. And then, data from the sensor is computer-processed to classify the object. This object classification system successfully manipulates a real-time object identification by utilizing an intrinsic material property: thermal conductivity.
|