Nao humanoid for physical therapy rehabilitation

Rehabilitation of patients in the medical field is an extensive area of study where patients are assisted in recovery of their body functions. Furthermore, this enables regaining daily routine functions, speech impairment or physical movement in the body. Depending on the inability, the process is designed for each patient by planning specific goals. This thesis will focus on physical rehabilitation and a system implemented to act as a physical therapist humanoid using Artificial Intelligence and Robotics. Physical rehabilitation deals with all movement injury in the body. This section of rehabilitation has been growing with the addition of computer technology in the past decade. Currently, robotics is used in physical therapy where it assists a part of the human body through direct body-part attachment and assistance. Thanks to the development of artificial intelligence, human interaction with robots is now possible and the best example of study that is being done is creating a social robot as a companion for the elderly to help them with loneliness and emotional well-being. However, this thesis will use deep learning methods to program a humanoid that can act as a customized physical therapist which will be implemented on a programmable humanoid robot, the Nao. This companion will show the movements associated with the recovery goals and provide direct feedback on the performance of the patient. Therefore, the design of this system will contain natural language processing and computer vision for human-based interaction and communication. Both these processes are neural networks that provide analysis and feedback. It will also outline and evaluate the methods that are used for this system to work efficiently and what kinds of changes can be made to improve the accuracy of the humanoid. In addition, a brief discussion will show the scalability and the future developments necessary to complete the system. The benefits of having a humanoid is that the patient will be able to perform his recovery at home at any time without travelling and will be able to customize the robot according to his or her process of recovery.