Development of an AI-based Wearable HumanComputer Interface for Assistive Computing

Patients with upper limb disabilities encounter major difficulties when they need to complete computer tasks because these tasks require special input devices, which can only be operated through precise small movements and ongoing body contact. The implementation of various proposed assistive technologies faces barriers because the systems require expensive components, operate through difficult designs, and lack portable functionality. This paper demonstrates an Artificial Intelligence (AI)-powered wearable human-computer interface, which enables users with upper body mobility restrictions to use computers through minimal body movements. The system uses a wearable sensor module, which connects to an ESP32 microcontroller for wireless communication through Bluetooth Low Energy to operate as a standard Human Interface Device that needs no extra drivers or software to function. The system uses a multi-tap input mechanism, which allows users to enter text quickly while experiencing less physical discomfort. The system includes an AI-based word prediction system, which offers context-based word suggestions to users, thus enabling them to type faster and with less effort. The system includes a graphical user interface that users can customize to receive visual information and create their own key mapping while managing the system. The experimental evaluation results show that the system enables users to type more efficiently because it requires less body movement while maintaining stable input accuracy, which proves that the system functions effectively as an economical portable assistive device for users in both low-resource and home settings.