Diseño e implementación de un prototipo robótico para supervisión y teleasistencia enfocado a adultos mayores que padecen de alzheimer.

Abstract

This research designed and implemented a robotic prototype with the objective of assisting an elderly person suffering of Alzheimer's disease at home. The basic needs of a person suffering from this disease were investigated, as well as the causes and factors that bring this disease, in addition to the selection and purchase of the best components that make up the prototype. The prototype involves: a facial recognition system using an LPBH algorithm, responsible for detecting and recognizing the responsible person and the elderly to allow them access to the pill dispenser, video calls and activities such as watching videos or listening to music, and a robotic system controlled remotely from the web and that has ultrasonic sensors for safety in the displacement. The pill dispenser is operated by stepper motors. A Raspberry PI 3 is used to control the facial recognition system, while an Arduino Mega is used for the robotic system. The prototype worked on average 3 hours 50 minutes of continuous time, the distance measurements of obstacle detection of the standard equipment and ultrasonic sensors are equal according to the inferential statistical test applied, also the estimated time of the robot to travel a certain distance is equal to the time measured in the practice of the movements. In addition, a percentage of success in facial recognition of 92% and the pillbox has delays in the last 2 days of testing because the internet network was used by other devices. It is recommended that the home network has total availability of internet service, otherwise, the person responsible for the elderly will not be able to remotely control the robotic prototype preventing its timely assistance to the elderly.