Two workers, same site, same heat. Only one of them is in trouble.
Site-level heat measures protect the average worker. Atsusa is an early-warning layer for the individual: it watches one person's heart rate against the real heat and humidity around them and warns them on their wrist while there is still time to stop, drink and cool down.
We have built it and tested the concept on one real case. We are now looking for one or two companies to help us find out whether it works where it actually matters.
No cost, no software to install on your systems, no commitment beyond the pilot. Atsusa is early-stage and not a medical device.
A site has one temperature. Your workers have twelve different tolerances.
WBGT thresholds, water-rest-shade cycles and supervisor training all work, and you should keep them. They describe the environment. They cannot tell you which specific person is already in trouble at 11:40 on a Tuesday.
New starters and people back from leave are at risk at lower workloads.
Cardiovascular reserve decides how long a body can keep shedding heat.
Both quietly reduce how much heat a worker can carry that day.
Common prescriptions blunt sweating and heart-rate response.
Protective clothing traps heat the ambient reading never sees.
Heat load carries over. Day three of a heatwave is not day one.
One worker, one wrist, one warning that arrives early.
The Apple Watch tracks heart rate continuously through the shift, along with how quickly it recovers between tasks. Where a live skin-temperature sensor is available, Atsusa uses that too.
Heat and humidity where the worker actually is, not a regional forecast. Humidity is what removes the body's ability to cool itself.
Those inputs are combined into one number: an estimated core temperature, updated minute by minute. It starts drifting upward before a worker feels unwell, which is what makes an early warning possible.
A haptic tap and a plain instruction: stop, get shade, drink, cool down. No app to open, no reading to interpret, nothing to remember.
- An individual early-warning layer on top of your existing programme
- A wearable-based signal that arrives before symptoms are obvious
- A worker-first alert: the person in the heat gets told first
- An early-stage product looking for real-world evidence
- A medical device or a diagnosis
- A clinically validated system
- A replacement for your heat plan, WBGT monitoring or supervision
- A worker-tracking or productivity tool
What the warning looks like across a working day.
Pick a working environment. The line is estimated core temperature across the shift: a modelled example built from published heat-physiology relationships, not measured pilot data. The point is the shape: the gap between carrying on and acting early.
Example profile: 38, moderately fit, second day back after time off
In this modelled example the worker has no clear symptoms yet. That is the whole point: the warning is designed to land while stopping for ten minutes is still an inconvenience rather than an incident.
The numbers behind this example
- Estimated core temperature at warning
- 38.3 °C / 100.9 °F(Modelled, not measured)
- Estimated peak if the worker continues
- 39.4 °C / 102.9 °F
- Heart rate at warning
- 148 bpm (≈82% of estimated max)
- Skin temperature
- 35.8 °C / 96.4 °F(Only available with a live skin-temperature sensor)
- Environmental conditions
- 34 °C / 93 °F air, 62% RH, WBGT ≈ 30 °C
- Workload estimate
- ≈ 5.5 MET, sustained
Modelled estimate. With an Apple Watch alone the core-temperature estimate carries roughly ±2.5 °C of uncertainty; adding a live skin-temperature sensor is expected to narrow that considerably. The direction and rate of change is the signal Atsusa uses, not an absolute number.
Want to try your own conditions? Run a full simulation with your own shift length, location and workload.
We have built something promising. We have not proven it yet.
If you are looking for a validated enterprise heat-safety platform, Atsusa is not it today. If you are willing to help find out whether an individual early-warning signal is worth having, read on.
One real-world case: the founder, at 40.7 °C core temperature after a half marathon, with medical staff present. That case is why Atsusa exists and it is also the honest limit of our evidence.
A working iPhone and Apple Watch app with continuous heart-rate monitoring, live local heat and humidity, tiered escalation and on-wrist haptic alerts. It runs today.
Whether the warning fires early enough, often enough, and rarely enough in a real occupational setting. Whether workers act on it. Whether supervisors find it useful. Nobody can answer that from a desk.
False alarms that workers learn to ignore. Missed cases. Wearables not surviving a real site. We would rather find that out with you than pretend it cannot happen.
Free, small, and designed so you can walk away.
We are looking for one or two companies, not twenty. That keeps it close enough that we can react to what you see.
- Atsusa on iPhone and Apple Watch for a small group of workers, free for the whole pilot
- Setup and onboarding run by us, including a short briefing for the workers involved
- A written summary at the end: what the system saw, when it warned, and what we would change
- Direct line to the founder throughout, not a support queue
- One crew or team working in genuine heat, typically 5 to 20 people
- Consent from the workers involved. This is voluntary, worker-first, and they can stop at any time
- A named contact who can answer questions during the pilot
- Honest feedback, including the parts we will not enjoy hearing
- 01Intro call
30 minutes. Your environment, your existing heat plan, whether a pilot makes sense at all.
- 02Scope it together
Which crew, which weeks, what would count as useful. Written down in plain language.
- 03Consent and setup
Workers briefed, devices configured, alerts explained. Nobody is enrolled without agreeing.
- 04Run the pilot
A few weeks of real shifts. We stay close and adjust as needed.
- 05Review honestly
What worked, what did not, and whether it is worth continuing. No pressure to buy anything.
Alerts go to the worker first. Participation is voluntary and withdrawable. We collect only what is needed to evaluate the warning signal, and we agree in writing beforehand what you see and what we see. Atsusa is not a productivity or location-monitoring tool and we will not turn it into one for a pilot.
Tell us where your people work in heat.
A few lines is enough to start. I read every enquiry myself and reply personally, usually within a day.