Robotic Arm Control Through the Use of Human Machine Interfaces and Brain Signals

Abstract

Applications of electroencephalography (EEG) on humans dates to the mid-1920s. During this time, time, scientists and physicists focused their studies on the brain functionality of patients suffering from seizures and epilepsy. Since that time, electroencephalography has expanded its uses in the medical field to diagnosing a wide variety of medical conditions, such as sleep disorders, comas, and migraines. This project’s goal was to develop an intelligence system that can both read and classify EEG signals from a subject when he/she performs a facial gesture. The intelligence system would be designed to send control signals to a robotic arm to perform specific tasks, such as making a fist or grabbing an object. Following data collection, the intelligence system was able to successfully differentiate the signals produced from the various facial gestures. The robotic arm was then programmed to perform a specific gesture associated with each facial gesture, based on the result produced by the intelligence system. Based on the success, additional facial gestures would be added to the intelligence system with the goal of making it more effective. With a larger database of facial gestures, the robotic arm would be able to perform more physical movements, ultimately providing a wider range of motions for users to select from. In the end, the robotic arm would eventually be applied in real-time scenarios to assist individuals who are paraplegics, those that suffer from nerve damage at amputation sites, and other disabilities that limit movement.