Design and Manufacturing of Cost-Effective Hand Exoskeleton for Paralysis Patient

The hand gestures required to pick up an object are predicted by sensors like Electromyography (EMG) in most of the hand exoskeletons. However, due to the sensor's noisy data, predicting hand gestures with it becomes challenging. Alternative hand gesture prediction technology, such as speech recognition, may be able to solve the aforementioned issue. An artificial intelligence technique called voice recognition translates speech input into written data. The objective of this research is to use voice recognition to create a low-cost hand exoskeleton for paralysed patients. The hand exoskeleton can provide shoulder, elbow, and finger movements for gripping an object. The user needs to speak the hand motion that he wants to employ to grab a certain object, and the hand exoskeleton takes it as a input. The microprocessor will direct the exoskeleton to perform the desired gesture once the gesture has been identified. The accuracy of anticipating a hand gesture using an EMG sensor and voice recognition are compared in the paper.