Abstract
This study assessed eight hand gestures using two MetaMotionR (MMR) environmental sensors. With one sensor placed on the top of the hand and the other on the wrist, this experiment aimed to determine the best sensor placement to control mobile robots. By analyzing the information in time and frequency and extracting time-frequency domain features performed gestures could be identified. Using an ensemble KNN, the system achieved 98.6% accuracy for the eight hand gesture recognition tasks. Therefore, it was determined that the best location for the sensor was on the combined locations of the top of the hand and wrist.