Even if someone falls alone at night,
staff will know in under a minute.
Whether they collapse at night, or fall in a hallway.
MoniCare sends alerts in seconds to eliminate delayed discovery, while care records reduce the burden of documenting your response.
At night, one caregiver may be responsible for more than 20 residents
As many as 71% of overnight shifts operate without a nurse. Across care facilities, residents who fall may remain undiscovered for hours.
operating without a nurse
one overnight caregiver
A fall can become a legal dispute
Courts have awarded more than KRW 200 million in fatal fall cases. Without objective evidence that staff responded immediately, the facility may bear the full burden of liability.
Share of patient-safety incidents caused by falls in long-term care hospitals
Falls reported by medical institutions in 2023
Potential damages in a fatal incident
Sensors and manually monitored CCTV
both miss the moments that matter
Every conventional approach leaves a critical gap. MoniCare uses AI to close those gaps.
Residents may refuse or lose them. They cannot detect a fall from bed or an incident while moving through a hallway.
They detect only where installed, leaving bathrooms and hallways as blind spots.
Useful only after the event. Overnight teams do not have enough staff to watch every screen in real time.
Contactless AI monitors around the clock, alerts staff as soon as it detects a fall, and creates the record automatically.
Overnight fall detected — 02:30
Unusual movement is detected in Room 302 on the third floor while the assigned caregiver is assisting another resident.
Camera detection
24/7 AI monitoring across rooms, hallways, and common areas
T+0 secAI analysis
Fall classification with eight algorithms and secondary VLM verification
T+3–5 secStaff notification
Simultaneous delivery to mobile, PC, display board, and LED
Within T+8 secOn-site response
Staff arrive and begin care; notification receipt time is logged automatically
T+1–3 minAutomatic record
Automatically store snapshots, VLM analysis, and the care timeline
Completed automaticallyMobile app
Receive alerts while moving through the facility
PC web dashboard
Nurses' station pop-up and audible alert
Display board
Large-screen alert in hallways and the nurses' station
LED indicator
A flashing light at the room entrance identifies the location instantly
Bed exit and unstable standing posture detected
AI analyzes skeletal coordinates to detect unstable standing, compares it with the resident's personal risk baseline, and determines that the risk level has risen.
Fall confirmed and VLM analysis begins immediately
After the primary algorithm detects the event, a vision-language model rechecks the snapshot and produces a fall probability, severity assessment, and recommended action.
Staff alerted across four channels at once
The assigned caregiver's phone, nurses' station PC, hallway display, and LED indicator activate simultaneously—within eight seconds of detection.
Staff arrive and begin care
Arrival and treatment start times are logged automatically, completing the timeline from alert receipt to on-site response.
Care record and response evidence created automatically
Before-and-after snapshots, VLM analysis, alert history, and care records are stored together in one audit report.
Delayed discovery — potentially hours
The fall may go undiscovered until the next round, increasing the risk of fracture or cerebral hemorrhage. With no record from the time of the incident, the facility cannot demonstrate its response.
Detection → Response — within 3 minutes
An alert is sent immediately, staff arrive within three minutes, and every step is timestamped automatically.
Before MoniCare
Falls detected only during overnight rounds
A fall during rounds made every 30 minutes to 1 hour cannot be discovered until the next round
Reliance on verbal and handwritten handoffs
Overnight events can be omitted during handoff or documented from memory
Care records written manually after the event
Records reconstructed from memory may contain inconsistent times and details, weakening evidence of the response
No objective evidence in a dispute
Without objective proof of immediate action, the facility may bear the liability
With MoniCare
Continuous 24/7 AI monitoring
AI analyses whenever the camera is on, with no overnight gaps between rounds
Automatic alerts + live event log
Alert delivery, receipt, and confirmation are stored automatically with timestamps accurate to the second
Automatic care records
Detection, alert, response, and care are linked automatically in one record
Instant response-evidence package
Export snapshots, VLM analysis, the event timeline, and care records in one report
Unified resident profile
Room 302, 3F · 3 connected cameras
Snapshots
Image sequence before and after the incident
VLM analysis
Evidence for severity and recommended action
Timeline
Events recorded to the second
MoniCare automatically packages material for municipal inspections, family disputes, and legal proceedings. Snapshot images, VLM analysis, a second-by-second timeline, and care records are exported in one PDF.
Dual-verification detection accuracy
Site-specific accuracy targets supported by a primary AI algorithm and secondary VLM check
Combined detection algorithms
Analyse posture, time series, bed occupancy, and personal baselines together to reduce false positives
Individual resident tracking
Recognize the same person across cameras and track risk through hallways and other rooms
Data sent outside
On-device AI keeps personal information from being sent to external servers
Site consultation
Assess the layout, camera positions, and implementation scope
2–5 daysCamera installation
Connect existing CCTV or install new cameras
1–2 weeksAI setup and calibration
Optimize personal baselines and alert thresholds for the site
1–2 weeksStaff training
Train staff on the dashboard and alert-response procedures
2–3 daysOperations and maintenance
Regular updates, remote support, and retraining
OngoingMoniCare has been deployed in long-term care hospitals, nursing homes, rehabilitation centres, day and night care centres, and municipal facilities. Both on-premises and cloud deployment are supported, so each facility can choose the model that fits its security policy and network environment.
After an on-site review of camera resolution, angles, and placement, we confirm which cameras can be integrated.
We recommend new cameras only where integration is not feasible.
A dual-verification process combines primary AI detection with a secondary vision-language model check to minimize false alerts.
MoniCare also learns a personal baseline for each resident and responds only to movement that departs from the norm.
During rollout, we tune alert thresholds to the facility's operating conditions.
Original video is never sent to an external cloud server.
Snapshots and record data remain inside the facility, with role-based access separated across viewer, nurse, admin, and super roles.
Assigned caregivers only need to receive the mobile alert and tap to confirm it.
The PC web dashboard runs in a browser with no separate installation.
Training materials and on-site manuals are designed around shift-based work.
It can identify unsafe walking posture in rehabilitation centres and wandering or stair-access risks in day and night care centres.
Detection categories are tailored to each facility type.
MoniCare's care records are designed from the outset for audit and response documentation.
An immutable audit trail of edits and cancellations, second-level timestamps, snapshot sequences, and VLM evidence can be exported together as one response-evidence package.
Report formats tailored to municipal inspection requirements are also supported.
Build the safety system that best fits your facility.
This product is not a medical device and does not replace medical diagnosis or prescription. Fall-detection accuracy may vary depending on the environment and installation conditions. Statistics and figures follow the source-attribution standards in the supplied documentation.