chapter  11
28 Pages

Human Peripheral Nervous System Controlling Robots

WithPANAGIOTIS K. ARTEMIADIS, KOSTAS J. KYRIAKOPOULOS

Contents 11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276 11.2 EMG-Based Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278

11.2.1 System Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278 11.2.2 Background and Problem Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 11.2.3 Recording Arm Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279 11.2.4 Recording Muscles Activation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 11.2.5 Dimensionality Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281 11.2.6 Motion Decoding Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286 11.2.7 Robot Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288

11.3 Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 11.3.1 Hardware and Experiment Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 289 11.3.2 Method Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290 11.3.3 EMG-Based Control vs. Motion-Tracking Systems . . . . . . . . . . . 294

11.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296 Appendix 11.A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297

11.A.1 Arm Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 297 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300

With the introduction of robots in everyday life, especially in developing services for people with special needs (i.e., elderly or impaired persons), there is a strong necessity of simple and natural control interfaces for robots. In particular, an interface that would allow the user to control a robot in a continuous way by performing natural motions with his/her arm would be very effective. In this chapter, electromyographic (EMG) signals from muscles of the human upper limb are used as the control interface between the user and a robot. EMG signals are recorded using surface EMG electrodes placed on the user’s skin, letting the user’s upper limb free of bulky interface sensors or machinery usually found in conventional teleoperation systems. The muscle synergies during motion are extracted using a dimensionality reduction technique. Moreover, the arm motion is represented into a low-dimensional manifold, from which motion primitives are extracted. Then using a decoding method, EMG signals are transformed into a continuous representation of arm motions. The accuracy of the method is assessed through real-time experiments including arm motions in three-dimensional (3D) space with various hand speeds. The method is also compared with other decoding algorithms, while its efficiency in controlling a robot arm is compared to that of a magnetic position tracking system.