Preliminary Fall Detection/Prediction Data Science Project

Fall detection systems, are targeted to older adults living alone to identify fall events and mitigate prolonged wait times to treat injuries typically. These systems use a wearable, such as a Ffitbit or Apple watch, paired with an algorithm to detect falls and alert caregivers or emergency services. However, large variability in type and circumstances of falls (e.g., from a small height with low impact) are a problem when trying to accurately detect and predict falls. The proposed approach aims to explore and exploit data gathered from NurtureWatch alarm to improve fall detection algorithms through the use of machine learning techniques. With the addition of machine learning Jabber Monkey can provide increased service and support to its clients in the event of a serious injury from a fall.

Faculty Supervisor:

James Tung

Student:

Mina Nouredanesh

Partner:

MoviWear

Discipline:

Engineering - mechanical

Sector:

Life sciences

University:

University of Waterloo