Regresar

Indoor Smart Walker Based on Artificial Vision for the Autonomous Movement of Seniors with Mobility Problems and Reduced Vision

Abstract:

The usual solution to mobility problems for the elderly lies in the use of conventional canes and walkers. However, when the older adult also has vision problems, these devices are not sufficient to provide safe walking, and the senior suffers from constant falls. This article proposes a smart walker based on artificial vision for indoor use in nursing homes that safely enables autonomous mobility for visually impaired seniors. The walker has been equipped with activation switches on the handlebars, ultrasonic sensors, a Kinect visual sensor, traction motors, and a controller. The control algorithm is based on reactive navigation, odometry and artificial vision and is designed to avoid obstacles and detect dangerous situations such as stairs. The control unit is a Raspberry Pi running Raspbian and relies on message passing between processes and visualization. The prototype of the smart walker is implemented in the nursing home “Luis Maldonado Tamayo” (Pujilí-Ecuador), where tests are being carried out with elderly persons who still have motor skills and some partial vision impairment. The walker can effectively detect obstacles in a distance range from 50 – 86 cm. Almost 90% of small, 96% medium, and 98% dynamics obstacles are avoided. The demonstration video is available at https://youtu.be/MJzvzIily4k.