The Kepler night nurse software
Our AI software enables healthcare facilities to reduce their night staff from three caregivers to one caregiver for every 100 clients. The software reduces false alarms by 99%. This helps prevent staff overload and thus reduces stress.
In addition to fall detection, the technology has also integrated lying position detection. This technology gives caregivers crucial information about clients at increased risk of pressure ulcers.
Trusted by international healthcare organizations and hospitals
Our fall detection and fall prevention software
The Kepler Vision state-of-the-art AI software is deployed in healthcare facilities. The software alerts and articulates unsafe situations in the client’s room with unprecedented reliability.
Our fall detection and fall prevention software
The Kepler Vision state-of-the-art AI software is deployed in healthcare facilities. The software alerts and articulates unsafe situations in the client’s room with unprecedented reliability.
Kepler Night Nurse
Using Kepler Night Nurse will provide your organisation with the following benefits, among others:
Helps prevent staff shortages
Less falls
Integrated with your NCS
How Kepler Night Nurse works
Kepler Night Nurse is an AI-powered software application that enables optical sensors to detect falls, unsafe situations, and nighttime activity in patient and resident rooms.The system continuously understands human posture and behavior without recording video, ensuring both safety and privacy.
-
An optical sensor is mounted on ceiling with a clear view of the bed and surrounding area.
If the optical sensor is “dumb,” an additional server is installed running the Kepler Night Nurse software. One Edge Appliance serves up to 50 optical sensors.
If the optical sensor is “smart,” such as the Mobotix C71, the software runs embedded on the sensor. No server is needed.
What happens then:
- The sensor captures a live video stream.
- No video is recorded or stored.
- In the case of a smart sensor, all processing takes place on the camera. In the case of a dumb sensor, processing occurs on the on-premises Edge Appliance.
-
If an event is detected and care workers need to be notified, Kepler Night Nurse sends an event notification to a nurse call system. The nurse call system will, in turn, route the message to the handheld devices carried by the care location staff.
When using a modern REST-style interface, the integration is very straightforward. The Kepler Night Nurse has already been integrated with dozens of nurse call systems.
Supported connection protocols
The Edge Appliance on which Kepler Night Nurse runs in a customer network typically has a direct TCP/IP connection to the nurse-call system (though it’s also possible to route this connection over the public Internet or through secure tunnels). This section contains an overview of the supported connection protocols.
REST
REST over HTTP(S) is the preferred way to connect Kepler Night Nurse to a nurse call system.
Custom
Other connection protocols are possible but may require more development time and effort. An additional risk is that these protocols do not offer the same level of reliability and security as REST.
Taking the above into consideration, examples of custom protocols that can be supported: network sockets, queues, and e-mail.
Supported authentication methods
Connecting to the nurse call system can be done with authentication or without authentication, depending on the customer network situation and requirements. HTTP tokens, HTTP Basic Auth, and custom authentication mechanisms are possible, although the latter require more development time and effort.
REST endpoint interface
For the preferred REST connection protocol, the Kepler Night Nurse expects a POST endpoint on the nurse-call system to deliver notification events.
Failure handling
When a notification fails to deliver via the REST protocol, the Kepler Night Nurse will retry two more times with a 2-second delay between each.
-
- During installation, the system integrator configures the room from a technical perspective. This merely means that the location of the exit door and the bathroom door is specified. This way, the Kepler Night Nurse can alert if a resident spends too long in the bathroom or leaves their room.
- Using the nurse call system, the caregiver selects which alert to activate for what room, such as fall, out-of-bed, sitting-on-bed-edge, and so on.
-
The Kepler Night Nurse AI understands human shape and movement, much as a caregiver observes a resident.
The system recognizes, amongst others:
- Lying, sitting, standing, walking
- Sitting on the bed edge
- Leaving the bed or the intention to leave the bed
- Slow falls and hard falls
- Presence in bed (any position)
-
Based on real-time analysis, the Kepler Night Nurse AI decides whether a situation is normal or requires attention.
For example:
- A resident turning in bed → No alert
- A resident’s pillow falls on the floor → No alert
- Resident sitting on the bed edge at night → Immediate alert
- Resident collapsing or falling → Immediate alert
This reduces false alarms and increases the resident’s safety.
-
If the Kepler Night Nurse detects a concerning situation, it sends an alert through your existing communication workflow.
Supported integrations:
- Smartphones
- DECT phones
- Smart pagers
- Nurse call systems
- Central monitoring dashboards
Alerts can include a privacy-friendly “face blurred” visualization so caregivers can see what is happening without exposing identity. However, because the alerts are very reliable, there is no need to check visually.
-
The early detection helps staff: ·
- Prevent falls by responding before the resident gets out of bed
- Assist immediately after a fall
- Reduce unnecessary room checks
- Spend more time on rooms where care is truly needed
This is especially valuable at night when staffing is limited.
-
The system improves over time through:
- Updated detection models and AI software
- New behavior recognition (e.g., wandering detection, getting-up-from-chair detection, intruder detection, prolonged lying position detection)
Facilities often experience that their staff need to walk less, receive fewer false alarms, and spend far more efficiently on nighttime care after installation.
Compared to other systems, Kepler Night Nurse is very accurate and only reports something when something happens. Below is a list of the notifications that can be sent. These notifications can be configured from a self-service portal when and if they should be shown.
Man down
Sitting on the edge of the bed
In bed
Sitting position in bed
Out of bed
Out of room
In bathroom
Getting up from chair
Sitting on the floor
Lying position
Uniform detection
Advantages Kepler Night Nurse
Using Kepler Night Nurse provides your healthcare organisation with many benefits:
Helps prevent staff shortages
Less falls
More safety
How Kepler Night Nurse works
Kepler Night Nurse is an AI-powered software application that enables optical sensors to detect falls, unsafe situations, and nighttime activity in patient and resident rooms.The system continuously understands human posture and behavior without recording video, ensuring both safety and privacy.
-
An optical sensor is mounted on ceiling with a clear view of the bed and surrounding area.
If the optical sensor is “dumb,” an additional server is installed running the Kepler Night Nurse software. One Edge Appliance serves up to 50 optical sensors.
If the optical sensor is “smart,” such as the Mobotix C71, the software runs embedded on the sensor. No server is needed.
What happens then:
- The sensor captures a live video stream.
- No video is recorded or stored.
- In the case of a smart sensor, all processing takes place on the camera. In the case of a dumb sensor, processing occurs on the on-premises Edge Appliance.
-
If an event is detected and care workers need to be notified, Kepler Night Nurse sends an event notification to a nurse call system. The nurse call system will, in turn, route the message to the handheld devices carried by the care location staff.
When using a modern REST-style interface, the integration is very straightforward. The Kepler Night Nurse has already been integrated with dozens of nurse call systems.
Supported connection protocols
The Edge Appliance on which Kepler Night Nurse runs in a customer network typically has a direct TCP/IP connection to the nurse-call system (though it’s also possible to route this connection over the public Internet or through secure tunnels). This section contains an overview of the supported connection protocols.
REST
REST over HTTP(S) is the preferred way to connect Kepler Night Nurse to a nurse call system.
Custom
Other connection protocols are possible but may require more development time and effort. An additional risk is that these protocols do not offer the same level of reliability and security as REST.
Taking the above into consideration, examples of custom protocols that can be supported: network sockets, queues, and e-mail.
Supported authentication methods
Connecting to the nurse call system can be done with authentication or without authentication, depending on the customer network situation and requirements. HTTP tokens, HTTP Basic Auth, and custom authentication mechanisms are possible, although the latter require more development time and effort.
REST endpoint interface
For the preferred REST connection protocol, the Kepler Night Nurse expects a POST endpoint on the nurse-call system to deliver notification events.
Failure handling
When a notification fails to deliver via the REST protocol, the Kepler Night Nurse will retry two more times with a 2-second delay between each.
-
- During installation, the system integrator configures the room from a technical perspective. This merely means that the location of the exit door and the bathroom door is specified. This way, the Kepler Night Nurse can alert if a resident spends too long in the bathroom or leaves their room.
- Using the nurse call system, the caregiver selects which alert to activate for what room, such as fall, out-of-bed, sitting-on-bed-edge, and so on.
-
The Kepler Night Nurse AI understands human shape and movement, much as a caregiver observes a resident.
The system recognizes, amongst others:
- Lying, sitting, standing, walking
- Sitting on the bed edge
- Leaving the bed or the intention to leave the bed
- Slow falls and hard falls
- Presence in bed (any position)
-
Based on real-time analysis, the Kepler Night Nurse AI decides whether a situation is normal or requires attention.
For example:
- A resident turning in bed → No alert
- A resident’s pillow falls on the floor → No alert
- Resident sitting on the bed edge at night → Immediate alert
- Resident collapsing or falling → Immediate alert
This reduces false alarms and increases the resident’s safety.
-
If the Kepler Night Nurse detects a concerning situation, it sends an alert through your existing communication workflow.
Supported integrations:
- Smartphones
- DECT phones
- Smart pagers
- Nurse call systems
- Central monitoring dashboards
Alerts can include a privacy-friendly “face blurred” visualization so caregivers can see what is happening without exposing identity. However, because the alerts are very reliable, there is no need to check visually.
-
The early detection helps staff: ·
- Prevent falls by responding before the resident gets out of bed
- Assist immediately after a fall
- Reduce unnecessary room checks
- Spend more time on rooms where care is truly needed
This is especially valuable at night when staffing is limited.
-
The system improves over time through:
- Updated detection models and AI software
- New behavior recognition (e.g., wandering detection, getting-up-from-chair detection, intruder detection, prolonged lying position detection)
Facilities often experience that their staff need to walk less, receive fewer false alarms, and spend far more efficiently on nighttime care after installation.
Compared to other systems, Kepler Night Nurse is very accurate and only reports something when something happens. Below is a list of the notifications that can be sent. These notifications can be configured from a self-service portal when and if they should be shown.
Man down
Sitting on the edge of the bed
In bed
Sitting position in bed
Out of bed
Out of room
In bathroom
Getting up from chair
Sitting on the floor
Lying position
Uniform detection
Kepler Night Nurse in comparison to other systems
-
The system recognizes unsafe situations within seconds and immediately alerts care staff.
-
No. Monitoring is completely passive and does not require wearables or buttons.
-
It recognizes early-risk behaviors such as sitting on the bed edge or attempting to stand up.