29. Mind Controlled Drones: Training EEG Signals to Control Drone Flights

Abstract

We developed a training system that transformed signals from a 14-channel EEG into commands to fly a remotely controlled drone via a heads-up display. The goal of this project was to create an efficient training program for use with affordable EEG headsets (Emotiv, Inc. 2014), creating reliable signals to drive remote controls for various devices, over wi-fi networks. The training program used standard software (e.g. The Emotiv EPOC Control Panel and the Emotiv Cognitiv Trainer) to strengthen signals through practice during multiple one-hour training sessions. After the signals were reinforced successfully, practice sessions with these thoughts are then used to drive the movement of a virtual cube (Emotiv, Inc. 2014). The cube movements simulated flight in a simplified environment. These biofeedback programs and our training directions should allow for home training and successful deployment of the EEG headset. The signal strength was measured by the number of trials it takes for the trainee to complete directed movement to criterion (8/10 error free trials) simulating flight. Each movement (push, lift, right, and left) was trained to criterion in a few sessions (M=5, SD=2). Subjects were able to complete each movement at a high percentage of success (M=73.69%, SD=7.76%). The training period was approximately five hours for each subject to run through every step of our program; including the set up of the EEG, creating the initial signals, and the reinforcement of the commands to strengthen the brain waves.