A research & engineering project

Warning workers
before the heat
becomes deadly.

Saya is a low-cost wrist wearable that detects dangerous heat strain in outdoor workers — and warns them early, before it turns into heat stroke. Built from the ground up: hardware, firmware, and a field study in Kuwait's 46 °C summer.

46°C
Kuwait summer highs workers labor through
3
Vital + environmental signals fused for early warning
<$70
Target cost — built to be affordable at scale
25–30
Workers in the planned field study

The why

It started at a dinner table.

I was twelve, stranded in Qatar during the pandemic, watching a country build World Cup stadiums in heat that could kill the people building them. We talked about it over dinner — the reforms, the scrutiny, and the workers who didn't come home. Many had crossed the Gulf from the subcontinent, the way my own great-grandfather once did, looking for work.

For years that feeling just sat there, useless. Empathy without a tool is only a feeling. Saya is my attempt to give it a tool — something that watches for danger a worker can't feel coming, and speaks up while there's still time to act.

— Veer Sahni

How it works

It feels what the body feels.

Saya continuously reads a worker's vital signs and their immediate environment, and escalates a warning through three channels you can't miss — even in blinding sun and site noise.

Senses strain early

Heart rate, skin temperature, sweat response, and ambient heat & humidity — watched continuously, so rising strain is caught on the way up, not after collapse.

Warns unmistakably

The whole face glows — visible across a worksite in direct sun. A 90 dB alarm and a wrist buzz make sure the worker, and those nearby, know to stop.

Remembers everything

Every reading is timestamped and stored, so the data can answer a hard question: does danger really keep to the hours the law assumes it does?

Watch

The build, and the field.

A short film covering how Saya was designed and built — from breadboard to manufactured circuit board — and the field trials with outdoor workers in Kuwait.

Coming soon

Filming during the August–September field trials. The video will be embedded here.

Inside the watch

Every part, in its place.

A working 3D model of the enclosure, dimensioned to the real 41.89 × 34.43 mm circuit board. Drag to rotate, scroll to zoom. The case is rendered transparent so the internal stack is visible — the board, the battery offset to one end, the haptic motors, and the clear polycarbonate windows the sensors read through.

Component positions are taken from the manufactured board's pick-and-place file, so the sensor windows sit directly beneath the sensors that use them. The heart-rate sensor fires down through the caseback window into the wrist, which is why nothing is allowed to sit in that optical path.

Build log

Building it in the open.

The real process — the breadboards, the schematics, the things that broke and what I learned. New entries as the project moves toward the summer field study.

June 2026 · this week

The advanced prototype comes together

Assembling the full sensor stack on a breadboard — heart rate, skin temperature, motion, ambient, and the real-time clock all talking on one bus. The messy, wire-everywhere stage where an idea becomes something that actually reads a pulse.

▶ Video / photos Drop today's breadboard build here
prototypesensorsESP32
June 2026

The circuit board takes shape

Turning the prototype into a real, wearable-sized board — a 4-layer design that fits everything into a watch. Working through the schematic with an engineer, learning how a tangle of breadboard wires becomes a clean, manufacturable layout.

▶ Schematic / screenshots Add the schematic view here
PCB designschematic4-layer
Spring 2026

From a question to a device

The project began as a research question — does heat danger really follow the clock the law assumes? — and grew into designing a device to actually measure it. Choosing sensors, defining the form factor, and learning what it takes to build hardware as a student.

▶ Early concepts Add design concepts / sketches here
researchdesignorigins

The research

A question the law may be getting wrong.

Many Gulf countries protect outdoor workers with a midday work ban — no labor during fixed afternoon hours in summer. It assumes danger keeps office hours. Saya's study asks whether that's true.

Aim 1 — Can it detect strain?

Evaluate whether a low-cost wearable can detect rising heat strain in real outdoor working conditions and issue staged early warnings.

Aim 2 — When does strain happen?

Examine whether dangerous strain occurs outside the legally protected midday hours — testing whether time-based protection matches when danger actually strikes.

The goal — better protection

If danger doesn't follow the clock, the evidence could help shape protection that follows the worker instead of the schedule.

Live companion tool

See the gap for yourself, right now

I built a live tool that pulls Kuwait's current temperature and humidity, computes the heat index using the same equation that runs inside Saya, and shows hour by hour how much of today's dangerous heat falls outside the protected midday window. No trial data needed — the public weather record already makes the argument.

Open the live tool →