Case study · Assisted living & memory care
Teton at Aquinas
How Aquinas Senior Living made Heritage Springs Montoursville the first US community to deploy anonymized AI monitoring at the point of care, and what 5 months of before-and-after data revealed.
The short version
A pioneering deployment, measured honestly.
Aquinas Senior Living partnered with Teton to deploy anonymized computer-vision monitoring at Heritage Springs Montoursville, the first US community to install the platform. Aquinas committed to a novel AI system in a live operating environment well before Teton was a known provider in the US market.
Ceiling-mounted sensors process resident movement, position and behavior on-device, without capturing identifiable imagery, preserving dignity while providing 24/7 passive observation across 57 resident rooms. The system surfaces real-time fall alerts, behavioral analytics, and proactive health signals, and includes Samwise, a generative AI assistant staff query in natural language.
After a stabilization period, the program ran a structured 1-month baseline followed by a 4-month active-use phase, enabling clean before-and-after measurement against the same residents and staff.
Operator
Aquinas Senior Living
Community
Heritage Springs Montoursville
Setting
Personal & memory care
Rooms observed
57 resident rooms
Measurement
5-month before/after
90%
Faster emergency reaction time
84%
Less time on the floor after a fall
89%
Of falls answered under 5 minutes
3hrs
App engagement / resident / day
The deployment
A novel system, switched on in a live care environment.
Being first in the US required Aquinas to commit to the technology before any peer community had validated it domestically. That openness is itself a creative act in a sector where AI adoption typically follows rather than leads.
Heritage Springs Montoursville had no domestic playbook. Workflow integration, alert configuration and staff training were developed in partnership and iterated in real operating conditions. The team did not wait for someone else to prove the playbook; they helped write it.
Real-time alerts
Falls and other risk events surface to staff devices the moment they happen, with in-room context before entering.
Behavioral analytics
Respiration, sleep and bathroom-visit patterns flag emerging conditions before symptoms are obvious.
Samwise, the AI assistant
A generative assistant staff query in natural language, surfacing insights they might not think to ask for.
57
resident rooms under continuous, anonymized observation. Vision data is processed locally and converted to anonymized representations before it ever leaves the room.
Emergency response
When help arrives in minutes, recovery is safer.
A caregiver arriving within minutes can assess for pain or injury before any movement, where getting up unaided risks a secondary fall. Self-recovery has dropped to under a third of falls.
90%
faster emergency reaction time after a fall
84%
less time on the floor before assistance arrives
89%
of falls now answered in under five minutes
<3min
average time to a caregiver reaching the resident
Prompt arrival also sharpens incident review: staff observe the undisturbed scene and pair it with the anonymized clip for a more reliable account than an after-the-fact reconstruction, turning every event into a learning artifact.
“
Teton gives our team insight into changes in our residents’ health, sleep, and fall risk, especially in memory care, where residents cannot express those changes themselves.
Emily Anthony
Personal Care Home Administrator, Aquinas Senior Living
AI adoption
A daily decision-support tool, not a background alarm.
Most AI deployments in long-term care are framed around a single output: a fall alert. Heritage Springs is using the same underlying AI more broadly. Staff query Samwise in natural language rather than navigating dashboards, and sustained usage shows the AI is being used as a daily tool by frontline caregivers.
Engagement changed the texture of care itself: visits got longer and more purposeful, not shorter.
3 hrs
active app engagement per resident, per day
+34%
average visit duration during the use phase
+17%
total staff time per resident
↓
lower staff turnover reported by Aquinas
Fall rate & review
Fewer falls, and far fewer unnecessary ER trips.
−9%
reduction in fall rate during the use phase, controlling for a single frequent-faller outlier
−40pp
reduction in ER send-out rate. Anonymized clips let staff see a controlled, non-injurious fall instead of sending out by default
The same clips surface root causes such as environmental hazards, footwear and gait changes, and serve as training artifacts to prevent the next fall.
Proactive health · Example early-warning signals
Respiration
The early infection.
Respiration tracking flagged potential concerns earlier during flu season, enabling closer monitoring of at-risk residents.
Sleep
The wellbeing barometer.
Too little sleep can signal agitation or cognitive change; too much can point to depression or other underlying conditions.
Bathroom visits
The quiet UTI.
A rise in frequency can be an early indicator of a UTI or other issue warranting further evaluation.
5
Five sub-scores combine into one Wellness Score, an at-a-glance read on each resident. By surfacing trends rather than isolated events, it helps the team notice when something is shifting, often before it escalates or a resident can put it into words.
Our approach & what’s next
No playbook, so the team helped write it.
Close collaboration, on-site from day one.
Being first meant no domestic reference existed, so Teton worked on-premise alongside Aquinas staff, tuning alert thresholds, escalation paths and training cadence to the community over the first months of use. That hands-on partnership is how adoption was earned: the AI became a tool that enhanced caregivers’ work rather than another system competing for attention.
Measuring honestly, even around an outlier.
A single frequent-faller disproportionately influenced the underlying picture. Rather than report the rosier number, Aquinas and Teton controlled for the outlier together and reached a shared view of true performance, the rigor that underwrote portfolio-wide expansion.
A second wave of AI
Choosing to be first again.
The success at Montoursville led Aquinas to expand the platform across its entire portfolio, where it has become part of standard care. Aquinas will also be the first operator to deploy Teton’s new AI-powered e-call system, pairing active call-button infrastructure with the passive monitoring already in place, with both signals arriving on one staff device through the same prioritization logic.
- Expanded portfolio-wide
- AI-powered e-call pilot
- Passive + active, one device
- First operator to deploy
See what Teton could do in your community
Prefer the polished PDF? Enter your email at the top of the page and we’ll send it straight to your inbox.
Book a demo