Fellowship Square Mesa cuts monthly falls 85% — from 20 to 3 — with Helpany AI motion detection

Fellowship Square Mesa, a 125-resident assisted living facility in metro Phoenix, averaged 20 falls per month before deploying AI-assisted fall prevention. Falls are among the most serious health risks for seniors, driving 800,000 U.S. hospitalizations annually and $31 billion in medical costs. Staff lacked real-time data to identify which residents were at highest risk on any given day or to intervene during high-risk overnight hours.

The facility deployed Paul, a radar-based AI motion detector by Helpany, installed on room ceilings. Using radar (no camera or audio), Paul continuously analyzes residents' stride lengths, instability, postures, and gaits to assess fall risk. It compiles daily movement-change reports for staff, identifies the 10 highest-risk residents each day, and sends real-time alerts when at-risk residents attempt to get out of bed overnight so staff can provide assistance.

Within six months of Paul's January 2024 deployment, monthly falls dropped from an average of 20 to approximately 3 — an 85% reduction — with zero overnight falls and no life-altering injuries recorded. The improvement has sustained for nearly two years. Staff also began using Paul's motion tracking data to detect residents in early stages of infections, enabling antibiotic treatment at home and avoiding prolonged hospital stays.

• AI-powered passive monitoring (radar, no camera) can achieve dramatic fall reductions while preserving resident privacy and dignity — a key barrier to adoption in senior care settings.
• The tool's value lies in surfacing actionable risk data to staff, not in autonomous intervention: "Paul is not preventing falls. Staff members are preventing the falls by having the right information."
• Continuous gait and movement analysis has secondary clinical value beyond fall prevention, including early detection of infections before they escalate to hospitalization.