Home Airflow Simulation and Cooling Optimization app/website.

Short version: totally buildable — and the scary-sounding part (3D modeling + airflow sim) is the easy part. Your phone can scan a room and auto-detect windows and doors, and pulling live wind for your city is trivial. The hard part is accuracy. Two reasons a phone-only version comes out plausible-but-not-quite-trustworthy: • The wind your weather app reports is measured out in the open (~10m up). By the time it reaches your windows it’s been bent — sometimes reversed — by neighboring houses, fences, and trees. • The app has no idea how warm it is inside, so whether opening up actually cools you is a guess — and that’s the whole ballgame (you only want windows open when it’s cooler out than in). The fix: make it optionally use two cheap sensors. **• Indoor temp/humidity sensor** (\~$15–20, or just read your smart thermostat) → the app *knows* it’s cooler outside instead of guessing, and won’t tell you to pull in muggy air. **• Outdoor wind sensor** → the actual breeze at your house, so “open these two windows” is a real call, not a coin flip. Still works with zero hardware (scan + weather data), just with more hedged advice. The sensors turn “probably worth opening up” into “open the north window now.” A few honest questions, since the answers change what’s worth building: • Would you actually be fine putting a temp sensor inside and a wind sensor outside? Or is phone-only a hard requirement? • Do you have a smart thermostat (Nest, ecobee)? Then indoor temp comes free. • Simple “open up / close up now” nudge, or true per-window guidance? The second is where the sensors earn their keep. On “does this exist already” — two sites market almost exactly this, but I couldn’t find either as an actual shipping app (no store listing, dead download buttons), so treat them as unproven concepts, not competition: crossbreeze.app and getventilate.app.

Excellent - all makes sense. I want to be straight about the "fluid dynamic simulation" part, because it hides the hardest, most expensive piece. Two very different things get called that: Full CFD: the impressive version that actually solves the swirling air and draws streamlines through each room (what SimScale/Autodesk do). It chops the house into a fine 3D grid and grinds through the Navier-Stokes equations; it does not run on a phone. You'd do it in the cloud, and that's the catch: every run costs real compute money, the wind shifts constantly, and every house differs, so you're facing an ongoing per-user cloud bill for a cheap app. That doesn't obviously pencil out. A multi-zone airflow network model: the pragmatic version. It treats each room as a pressure node and each open window/door as a flow path, then solves where the air actually moves given the wind pressure on each face of the house (plus warm air rising). It's the standard method engineers use for exactly this and it gives you your answer ("north wind hits this face, open these for the best cross-flow"), light enough to run on-device, free, in real time. So you can get the per-window/door guidance without real CFD and without a cloud bill. The one honest caveat: the model needs an estimate of how wind pressure lands on each face; approximated from the home's shape or calculated once per house at scan time and cached, so daily updates stay free. That's exactly where your outdoor wind sensor earns its keep, and where the app can learn your specific house over a season ("north wind + these open → living room actually dropped 2°C"). TL;DR: the "which windows for a north wind" brain is very doable on-device. True full CFD is cloud-only and someone pays per run, so the realistic build is a lightweight airflow model that gives the same actionable answer without lighting money on fire. What do you gauge the "market" for this might be? Or, I'll candidly ask - how useful is this to you and what would you pay for the app (as a consumer)?

Great reply!