It is an unfortunate fact of life that people often fall down. What’s even sadder is that sometimes, they never get back up.

Patient fall injuries has become major concern for hospitals- in New Zealand, up to 52 per cent of reported serious injuries are due to patients falling during admittance. While certain solutions already exist in the market to detect falls, none exists that actively prevents falls from happening in the first place.

An opportunity to research this problem was generously funded and supported by various stakeholders, including the Canterbury District Health Board, the Canterbury Medical Research Foundation, the University of Otago’s Medical School and the University of Canterbury’s Engineering department. As the principal researcher, I was tasked with developing a proof-of-concept prototype to prevent this problem.

 

I began by reviewing the problem, existing solutions and interviewing the doctors and nurses at the hospital. A range of possible technological solutions were shortlisted:

  • RFID
  • BLE (Bluetooth Low Energy) also called iBeacons
  • Computer Vision solutions
  • Lasers
  • Ultrasonic sensors
  • Force Sensors
  • Stretch sensors

After further consulting with some engineering professors we decided to explore the BLE (iBeacon) solution.

By using iBeacons as wearable sensors (on a patient’s ankles), we can measure the signal strength (RSSI value) to calculate a patient’s position on a bed. I made the prototype using BLE equipped Arduino boards, and wrote a simple android app as a control unit to read and display the RSSI signal of the two sensors.
Below are pictures of the development and testing stages. Due to the commercial sensitivity of the project the full results are not published, but feel free to contact me with questions.
A slide from my presentation highlighting the problem
The two wearable ibeacons are transmitting a BLE signal which is processed by the app on the phone, acting as the central unit
Testing the RSSI signal strength at different distances and with obstruction testing. Here I am using arduino boards with BLE modules