Shade + double roof
A reflective removable top, ventilated gap and shaded seated zone form the core. Measure roof response against a matched reference.
CORE PROTOTYPEVELLORE, TAMIL NADU / OPEN ENGINEERING
Good shade.
Honest evidence.
A small shelter with a bigger question: can we make outdoor rest spaces work better, using less?
SAAYA combines passive cooling, local sensor control and an evidence trail that knows when to say “we don’t know yet.”
Interactive demonstrator · physical testing pending
01 / TABLETOP CONCEPT
Rotate the model. Lift the roof. Find every component.
Passive firstShade, airflow, repairable parts.
Local by designControls keep working offline.
Evidence before claimsEvery result carries its limits.
02 / COOLPROOF LAB
Make it fail.
See what it tells you.
Replay a repeatable, synthetic bench scenario. These numbers illustrate system behaviour; they do not predict cooling in Vellore.
SIMULATED DATA
Choose a condition
60 synthetic samples · seed 1 · 1 s sample spacing · playback accelerated. Parameters lock when replay starts.
Ready. No experiment started.
INDOOR AIR—Temperature, not safety
RELATIVE HUMIDITY—Environmental measurement
FAN FEEDBACK—Command: —
ELECTRICAL POWER—Missing is not zero
Seated air speed
Intervention Reference
Water and sensor feedback appear during playback.
EVIDENCE RECORD
Awaiting a declared experiment
An invalid comparison produces no improvement claim.
PRESENT TO YOUR PANEL
Every component. Every connection. Explained.
Explore the IoT block diagram, inspect labelled CAD, and explain why sensors, controller, display and actuators are placed where they are.
Open the teaching walkthrough →03 / OLD IDEAS, MEASURABLE QUESTIONS
Learn from buildings.
Test the mechanisms.
Traditional architecture offers useful principles. Each one still has to earn its place in this climate and at this scale.
Jaali + wind paths
Interchangeable screens explore shading versus airflow resistance. A windcatcher is an experiment, not an assumed cooling gain.
REPLACEABLE TEST PARTSTerracotta + water
Evaporation depends on weather and creates cleaning demands. A separate non-potable module is deferred until local tests justify it.
DEFERRED / NOT AUTOMATICOpen the IoT architecture and local working flow
Sensors → validation → independent local controls → source-labelled telemetry → CoolProof. The public site runs synthetic replays. Real USB telemetry is available through the local bench API.
04 / RESEARCH REGISTER
Show the source.
Keep the caveat.
Eight dossier references plus targeted research on traditional mechanisms, prior art and Vellore climate. Access depth is recorded; a citation is not prototype validation.
Loading research register…
05 / FROM DRAWING TO BENCH
Buildable parts.
Visible unfinished work.
A fabrication candidate, not a certified product. Real hardware measurements are still required before cooling or safety claims.
Prototype package
- ESP32 firmware with local control, fan tach, sensor validation and latched help.
- FastAPI + SQLite telemetry, experiment records and CSV/JSON export.
- Parametric enclosure, mounting provisions, sensor shields, roof parts and screens.
- Bench protocol, calibration records, supplier references and assembly instructions.
AI / READ-ONLY BY DESIGN
Explain the evidence.
Never invent the result.
The optional local-model analyst receives a finished experiment and a curated source register. Its explanation cannot change controls or the recorded outcome.
No model is connected to this public demo. Deterministic evidence checks run without AI. Learned control is deferred until measured training and independent evaluation data exist.
Run the local bench API ↗Real device monitor
When this site is served by the local bench API, this panel reads actual USB-forwarded measurements. Public hosting never pretends to host the device service.
No physical device connected.
06 / COST AND RESOURCES
Less than an AC?
A target to verify.
Compare full installed costs, then compare the services honestly. An open shelter does not provide a room AC’s temperature control or dehumidification.
Loading cost register…
Solar is optional and separately costed. Instrument rental, transport and fabrication must be priced before procurement. No purchase has been made.
07 / WHAT THIS DOES NOT PROVE
A useful prototype
is honest about its limits.
Scale
Miniature convection and radiation do not transfer directly to a worker-sized shelter.
Measurement
Fan RPM is not seated air speed. Temperature and humidity alone cannot establish occupational heat safety.
Real use
Dust, rain, corrosion, noise, water hygiene, maintenance and access to breaks can outweigh a laboratory gain.
Evidence
Thermal benefits, worldwide novelty, structural adequacy and fabrication fit remain unproven until tested.