Robot control based on human motion analysis with IMU measurements

The measurement of the operator motions can lead to innovation in the field of human-robot collaboration. Motion capture data can be used for robot programming by demonstration
as well as controlling and adapting robot trajectories. To measure motion, IMUs represent an interesting alternative compared to other sensing technologies such as vision. In order
to illustrate this strategy, a demonstrator has been created to command a robot by human motion. Two IMU sensor modules are set on the human arm and forearm. The orientation of
the module with respect to the inertial frame is computed by the fusion of accelerometer, magnetometer and gyroscope data. This information coupled with a simple human arm kinematic
model is used to compute the wrist trajectory. The accuracy of this measurement has to be quantified. For that purpose, the estimated trajectory based on the IMU measurement is compared to the trajectory measured using a reference 3D optoelectronic motion analysis system (Codamotion, Charnwood Dynamics Ltd) available in the Laboratory of Human Motion Analysis of the University of Liège.