Design and Implementation of Deep Learning Assisted Smart Wheelchair

This research paper delves into the multifaceted world of smart electric wheelchairs, focusing on their advanced control modes, wireless communication protocols, and AI-based movement capabilities. The introduction sets the stage by emphasizing the potential advantages of these wheelchairs for individuals with mobility disabilities, highlighting their transformative impact on autonomy and daily life. The literature review explores technical challenges, including the absence of standardized communication protocols and the need for enhanced obstacle detection and navigation algorithms. Additionally, it discusses the integration of AI technologies, which hold promise for further enhancing these wheelchairs’ capabilities. The methodology section details the approaches employed for manual, remote, mobile application, and AI-based control modes. The AI-based movement mode is particularly intriguing, utilizing computer vision and image recognition to enable autonomous wheelchair following. Data preparation and dataset creation are explained, along with the YOLOv8 model architecture used for AI-based movement. The results and discussion section showcases successful implementation of control modes, with users effectively operating the wheelchairs. Additionally, the AI model demonstrates excellent performance in object detection and following tasks. In summary, this research paper provides a comprehensive exploration of smart electric wheelchairs’ features and capabilities, shedding light on their potential benefits for individuals with mobility impairments and offering insights into the technical and AI-driven advancements in this field.