Four features, built on 2,000+ analysed falls, that adapt to each resident every day. Early communities are seeing fewer notifications, more residents monitored, and fewer falls.
Book a demoTraditional fall alarms treat every movement the same. We analysed over 2,000 confirmed falls across the US, Denmark, the UK, and Switzerland to understand which movements actually matter — and built a system that acts on what we found.
Optimised Alarms helps care teams prevent falls with less noise, so the right alarm reaches the right person at the right time.
Set alarm sensitivity by fall risk level — Low, Medium, or High — in two clicks. Each level maps to ML-trained rules for which transitions to alert on, when, and after how long.
An ML model analyses 21 data points per resident every day — sleep, respiration, mobility, fall history — and suggests an updated risk level with one-click apply.
Detect when a walking aid or wheelchair has moved out of reach. Alert staff before the resident gets up without it — the single largest category of preventable falls.
Autopilot applies predicted risk levels to alarm presets automatically every night. Every resident's alarms reflect their current risk without anyone needing to remember to update them.
Alarm presets replace broad, one-size-fits-all alerts with targeted rules calibrated against 2,000+ real falls. When the alarm goes off, it means something.
Daily risk predictions surface changes in sleep, respiration, and mobility the moment they happen — not months later at the next care review.
Autopilot closes the loop between prediction and action. Risk levels update overnight, every night, without staff needing to check a single setting.
Every alarm rule in Optimised Alarms is calibrated against real data from over 2,000 confirmed falls across care settings in four countries. Getting out of bed is 2.5x more likely to result in a fall. Night-time transitions are 63% more likely to lead to a fall. Walking aid users are 3.2x more likely to fall — and in 67% of those falls, they weren't using their aid. Previous fall in last 3 months means 5x more likely to fall again. Wheelchair users getting out of bed face 5.3x fall risk.
Book a demoThe features that help care teams act on the right signals, every shift.
Staff select a risk level and walking aid status during admission. The system configures the right alarms automatically — no manual tuning needed.
Night staff receive targeted alerts for bed exits, walking aid movement, and high-risk transitions — without the noise of broad alarms waking the whole floor.
After a fall, the ML model recalculates risk and Autopilot adjusts alarm sensitivity overnight. The resident's protection increases immediately.
Managers review predicted risk levels, alarm history, and trend data to inform care plans and communicate clearly with families about what's being done.
We rolled out Optimised Alarms at [TBD] communities. Fewer alarms. More residents covered. Fewer falls. Faster responses. Not a trade-off — all four improved at once.
Read the full resultsStaff choose a fall risk level (Low, Medium, High) and a walking aid status. Each level maps to ML-trained transition rules. Smart 5-minute deduplication ensures the same transition can trigger an alarm at night and stay silent during the day.
An ML model analyses 21 data points per resident every day — fall history, sleep regularity, respiration changes, mobility patterns, walking aid usage. It produces a daily predicted risk score with a one-click Apply button.
The sensor detects when a walking aid or wheelchair has moved more than 1 metre from the bed. Staff receive an alert — repositioning takes 30 seconds. For wheelchair users, the system monitors bed-exit transitions (5.3x fall risk) and alerts staff when a wheelchair user begins getting out of bed.
Autopilot runs daily at 3:50 AM. It takes the ML-predicted fall risk level and applies it to alarm presets automatically. If sleep has fragmented, respiration has shifted, or a fall happened last week, alarm sensitivity increases overnight. Staff manual decisions are never overridden.
Standard alarm settings use static thresholds that don't change. Optimised Alarms calibrates every rule against 2,000+ real falls and adapts daily based on each resident's current risk profile — sleep, respiration, mobility, fall history.
No. Fall Risk Prediction suggests a risk level, but staff always approve changes with a one-click Apply button. Autopilot automates the application of staff-approved risk levels but never overrides manual decisions.
Yes. Optimised Alarms is a software update — no new sensors, no installation visit, no downtime. If you're already running Teton, you can enable it today.
No. Optimised Alarms is not a medical device under EU MDR or equivalent frameworks. It helps care staff configure and maintain alarm settings based on operational monitoring data. Clinical responsibility remains with the care team.
Staff select a risk level and walking aid status per resident — two clicks. Autopilot handles ongoing updates automatically. Most communities are fully configured within one shift.
The model analyses 21 data points per resident daily: fall history, sleep regularity, respiration rate, mobility patterns, walking aid usage, transition frequency, time of day patterns, and more. All data comes from the Teton platform — no external data sources.
Yes. Each feature works independently. You can use alarm presets and fall risk prediction without enabling Autopilot. Autopilot simply automates the last step.