Design and fabrication of a self-feeding robot for Parkinson’s patients and the elderly

Abstract

CareFeeder is a smart assistive device designed to aid Parkinson's patients and elderly individuals who struggle with feeding themselves. This machine comprises three main components: an index tray, a 2 Degrees of Freedom (DoF) grip arm equipped with a spoon, and a CareFeeder application for easy control. This device configuration not only ensures efficient manufacturing but also enables widespread availability at an affordable price. In addition to assisting patients in their daily activities, CareFeeder aims to enhance metabolism through personalized nutritional support. Utilizing machine learning to analyze patients' nutritional consumption, CareFeeder can provide direct diet recommendations instead of relying solely on doctor advice. Specifically, CareFeeder features a 2 DoF grip arm and an index tray. Its operation involves using the grip arm to pick up food from the tray and feed individuals with limited mobility, such as Parkinson's patients or the elderly. The device can operate autonomously and wirelessly transmit meal records and relevant information to a mobile device via a web server. Finally, CareFeeder's management tool analyzes the data based on nutritional standards by using the camera to detect the user's mouth movements and predict the number of calories in the dishes. From there we can provide the most appropriate amount of nutrition for the patient. The controller then adjusts the manipulator's movements and the tray's automation according to the patient's health situation and specific needs.