NeuroGrip: Mind-Powered Prosthetic Arm for Enhanced Mobility

Abstract—This paper presents a cost-effective, brain-controlled prosthetic arm system using EEG signals acquired from the Neuphony EEG headset. The objective is to restore limb func- tionality for amputees by interpreting brainwave patterns to control a 3D-printed prosthetic arm in real time. The system architecture includes signal acquisition, preprocessing, feature extraction, and classification using the XGBoost machine learning model. EEG signals are filtered and segmented before extracting relevant statistical features, which are then used to train the model to recognize specific hand gestures. These gestures are translated into motor commands using an ESP8266 microcon- troller to control high-torque servo motors in the prosthetic arm. Experimental results demonstrate a high classification accuracy and responsive gesture execution, validating the system’s practi- cality. This work contributes to the development of affordable, intelligent prosthetic solutions with real-time control, promoting better integration between neural signals and mechanical motion. Index Terms—EEG, Brain-Computer Interface (BCI), Pros- thetic Arm, XGBoost, Signal Processing, Machine Learning, Motor Imagery, Real-Time Control, Neuroprosthetics, Feature Extraction, ESP8266, IoT-Based Prosthesis