W.A.R. – Wheelchair Activity Recognition

There are approximately 131,800,000 (1.85% of world population) people in the world that have a physical disability and require a wheelchair. Musculoskeletal pain because of wheelchair use is very common amongst wheelchair users. Optimal adjustment of seating position may prevent pain and other deceases, monitoring the seating position is an important tool to improve the quality of life for those who need use a wheelchair for long term. This information is important for the user who can receive instructions to improve the positioning of the chair and for the orthopedist who can more easily customize the wheelchair adjustments. Physical activity is important for the health and well-being of wheelchair users, using a wheelchair as a means of exercising will make an important contribution to the health status of its users. The creation of an integrated system that monitors and records the activity of the wheelchair, senses the wheelchair to carry out the monitoring, understands where to place the sensory devices and creates an intelligent system that, through the information collected, is able to classify the different positions of the wheelchair is essential for promoting the health and well-being of wheelchair users. This work shows the development of an monitoring system for wheelchair, the results indicate that the applied technology proved to be suitable and applicable for any type of wheelchair, manual or electric. The collected data allowed the development of a robust and highly accurate classification model, making it possible to classify the position of the wheelchair in real time, allowing the issue of alerts to the user. A mobile application was developed that concentrates all the information, making it available intuitively in two dimensions: wheelchair user and orthopedic surgeon.