SpxTrm AI: Video Surveillance to Reduce Falls

Award Date: 2018

Project Lead(s): Navjot Virk, Manager, Research & Practice

Project Title: SpxTrm AI: Empowering Caregiver Abilities By Employing Computer Vision in Long Term Care


What is the issue?

In Canada, about 15 per cent of long-term care residents experience a fall in any given month. Most falls are not observed by staff, making them difficult to prevent.

What did we do?

We evaluated video surveillance technology made by Spxtrum AI in a long-term care setting. Our goal was to determine if the software would alert staff about residents at risk of falling, enabling them to intervene before falls occur.  We conducted stakeholder engagement, education, and training sessions. Staff, residents, and families provided feedback. Spxtrum AI provided data on how the innovation was used. Maintenance and IT staff provided input on the pros and cons of implementing the technology. We also conducted a privacy impact assessment.

What did we find?

The technology worked well to identify when a fall had occurred. However, it was unable to predict falls in advance of them happening. Therefore, while it has the potential to help seniors and care providers, the technology is not effective, as it is, to prevent falls.