本文透過 Kinect 擷取手套特徵物件的 3D 位移資訊，以及使用 LPMS-B 六軸加速度感測器即時感測手 腕當前三軸角度姿態，整合成終端效果器六軸資訊，對機器手臂執行遠端操作。其中手套的設計結合 Arduino 的開關控制，遠端控制機器手臂終端效果器上裝設的機器夾爪開闔 (抓取或放置物件)，發展出一 套不必事先教導，可供人類藉著移動自己的手掌來直接指揮六軸機器手臂執行六軸同動運動，完成對隨意 物件所需的操作。此技術適合應用在隨機使用機械手臂執行物件操作的場合，以直覺手勢直接引導的操作 方式，免除示教過程，創造時間節約、使用效率和競爭力，提高智慧自動化水平。相同技術也可用在創造 難以教導或需長時間才能完成的機器手臂困難運動軌跡，再重複實施在機器手臂後續的物件操作上。這對 需要複雜軌跡才能完成的物件組裝而言具高度應用價值。

In this paper, 3D displacement information of the featured glove is captured by Kinect and the Euler’s angles of the operator’s wrist are postured by LPMS-B 6-axis acceleration sensor (from L-P Research) immediately. After that, the 3D displacement information is integrated with Euler’s angles into the integral 6-axis information for defi ning the real-time end-effector position and posture in the robot arm tele-operation. Furthermore, the featured glove was equipped with an Arduino circuit for controlling the on-off operation (grasping or releasing object) of the gripper on the end-effector. This innovative system can be used to control the 6-axis robot arm to operate a randomlypositioned object without pre-teaching. This technique is particularly suitable for operating randomly-positioned objects by using multi-axis robot arm. The operation is intuitive, without the time-consuming teaching process, save time, and increase effi ciency. The new system can also be used in generating trajectories for some hard-to-teach or high time-consuming object operations. The trajectory produced by the hand posture operation can be subsequently used on repeated object operations. This is a valuable tool for some hard-to-teach assembly tasks.