IoT-EnabledHumanoidRoboticArmOperatedvia NeuralSignals andEye-Tracking withBlink-Based Morse Communication
Authors: Mr. Aayush Sharma [1]
V.K Krishna Menon College, Bhandup (East)
Guide: Mrs. Kalpana Bandebuche [2] Assistant Professor, V.K.K. Menon College, Bhandup (East)
Abstract :
Amputations and neurologic disorders considerably restrict individuals' interaction with the environment. As per the WHO (2021), more than 30 million individuals are livingwith limb loss, and arm amputations accountforthe most prevalent. Existing BCI humanoid arms are still mostly restricted to the lab; invasive BCIs need brain surgery, whereas non-invasiveEEGmethodstypically requiremultiple electrodes because of signal attenuation. This work suggests an effective EEG-controlled humanoid arm system by employing six specific electrodes from a 32-channel data set (10–20 standard).It consists of three phases:signal preprocessing (band-passfiltering,CSP, and CWT), feature extraction by a pretrained VGG1c network, and classification to control the robotic arm. A LabVIEW GUI is used to control the arm's kinematics and dynamics. Experimental testing scored S0.2% accuracy in classification, allowing for consistent and naturalistic arm control. These outcomes demonstrate the promise of non-invasive BCIs to improve independence and quality of life in amputees
Keywords: Brain–Computer Interface (BCI); Electroencephalography (EEG); Common Spatial PatternCSP, humanoid arm,AssistiveRobotics.
