Elmbrook Court cuts falls 50% and saves 9 care hours per night with Ally AI resident monitoring

Night-time care at Elmbrook Court's 74-bed memory care facility was highly resource-intensive, requiring up to five staff on the memory care floor each night. Traditional bed sensors triggered frequent false alerts from minor movements, waking residents with dementia unnecessarily and trapping staff in a cycle of reactive, disruptive checks that compromised both resident sleep quality and dignity.

Elmbrook Court trialled 11 Ally AI resident monitoring devices using acoustic and motion sensing across two memory care floors. A phased onboarding approach gradually built staff confidence, reducing scheduled hourly checks first to two-hourly, then four-hourly, and eventually eliminating routine checks entirely for some residents. The system also provides daily summary emails for management oversight and stores audio/motion data for post-incident review.

Falls on the memory care floor dropped by 50% without increasing staffing levels, and infections among Ally-monitored residents fell by more than 50% over six months due to improved sleep quality leading to better nutrition and hydration. Operational metrics showed 9 hours of care team time saved per night, a 64% reduction in disturbing checks, and a 5x return on investment in staff time. Nine out of eleven monitored residents maintained or gained weight.

• Phased adoption that preserves existing routines while demonstrating visible results is the most effective path to staff trust and cultural buy-in for AI monitoring tools.
• Improved sleep creates measurable cascading health benefits—better nutrition, fewer UTIs, greater daytime engagement—making night-time monitoring a whole-day care intervention.
• Continuous passive monitoring enables earlier detection of deteriorating health (e.g., pre-UTI restlessness visible 24–48 hours before diagnosis), shifting care from reactive to preventive.