On the device by default
Processing runs where the camera is. Images and sound do not need to leave the device, and a deployment that does need to share data sets its own rules for it.
The Human-State Layer
One layer, placed where a person is part of the system.
Operators, drivers, residents and patients sit inside systems that watch everything except them. The gap is the same in every field: nobody follows how the person's state is changing. EmoPulse fills it with one layer that runs on the cameras the system already has.
OPERATOR MONITORING · DRIVER MONITORING · CARE & SENIOR SAFETY · CLINICAL WORKFLOWS
01 · One Layer
The system sees the room. It misses the person in it.
A vehicle has sensors, a plant has dashboards, a ward has monitors. What is missing is a running picture of the person using them. EmoPulse adds that missing human-state layer.
We supply a layer rather than an application. It integrates into a partner's stack, hardware and security boundary. Technical documentation and delivery terms are shared with partners on request.
Independent core
Patented. No dependence on anyone else's proprietary model: open-source components underneath, our own layer on top.
Built to be integrated
Edge or server. Air-gapped or connected. The same core engine is configured for a deployment rather than rewritten for it.
02 · Human State Over Time
A person is a trend, not a snapshot
One reading says little. What matters is how a person's state moves away from their own normal over minutes, hours and days. EmoPulse follows that movement and hands it to the system as a signal it can use. Each deployment takes only what its purpose needs, with no extra hardware and no cloud dependency.
The mechanics are on the Under the hood page.
03 · Solutions
Four settings, one layer
The layer stays the same. What changes in each field is the question asked of it, and the evidence and regulation that come with that question.
Operator Monitoring
The task is already monitored. The operator is not. EmoPulse adds the missing side: how the person running the equipment is doing.
For people in control rooms, cockpits and on machines, fatigue, vigilance and workload are read on the device. No wearables, and no images leave it.
Driver Monitoring
The car knows the road and its own speed. It does not know how the driver is doing. EmoPulse gives the cabin camera that second job.
It works with a standard cabin camera and keeps all computation inside the vehicle.
Open Driver Safety MonitoringCare & Senior Safety
Alarms and buttons answer something that has already happened. We are exploring a gentler question: can slow changes in how a person is doing be seen while help is still easy to give?
Concept stage. It could help identify changes that may precede risky situations, rather than only analysing events afterwards. The comparison is a concept illustration with simulated values.
Open the fall comparison (concept)Clinical Workflows
A visit is a snapshot, and most of a patient's life happens between visits. We are exploring whether camera-derived observations between visits could give clinicians a longer view.
Presented to clinicians as structured observations. Research and clinical validation come before any medical use. Not a medical device.
Open Clinic Stack04 · Demonstrators
The layer at work
Each demo below is the same layer doing a different job. The outputs differ, the layer does not. All of them show what happens when a system learns about the person as well as the place.
Biometric Signature
The same layer used for identity: a behavioural signature that keeps updating on the device and only holds when several channels agree at the same time.
Proof-of-concept demonstrator only and not a certified authentication system.
Open Biometric SignatureLive Dashboard
The layer reading a person in real time from a standard camera, in the browser.
Open Live DashboardSystems are getting very good at understanding their surroundings. The person inside them is still mostly a blank. EmoPulse is building the layer that fills it.