OEM site screen · screening-grade
Factory-siting screen for a robot or humanoid OEM site team: SITE climate is public MODELED dry-bulb (a factory-floor labor/HVAC screen, not a data-center campus); PLANT acres, MW, and rail are operator CURATED and never invented.
ERA5 dry-bulb via Open-Meteo. Last ~365 days ending 5 UTC days ago. Not live weather, not wet-bulb, not free-cooling.
Typed by the OEM site team. Empty stays empty. Completeness is a fill count, not a score.
Shortlist compare · not a ranking · plant fields are whatever you typed for that pin
screen: plant-plausible only if acres ≥ 200 and target MW ≥ 50 and rail-logistics is answered and SITE hours exist. Climate-band ≥ 40% of resolved hours is a MODELED note, not a gate. Else incomplete if SITE hours missing or plant completeness < 3 (acres / MW / rail-class; unknown MW is incomplete, never a fail); constrained if SITE hours exist, completeness is 3, and a hunch fails. The 200 acre / 50 MW / rail-answered cutoffs are this page’s v1 screen (hunches), not a defended industrial standard.
Screening-grade. Not underwriting. Not a site-selection stamp. Not an interconnect-queue product. Not a factory design. SITE hours are ERA5 grid-cell reanalysis via Open-Meteo, not a weather station on the plant. Climate-band is a labor/HVAC dry-bulb screen — not free-cooling for IT, not wet-bulb. Flood is a public FEMA NFHL pin query — not a survey, not a pad study, not FIT. PLANT numbers are whatever the operator typed — Mill does not fetch CoStar, LoopNet, queues, or any incentive database. The FIT cuts are v1 hunches shown on this page, not industry fact. On mill.tools, ERA5 is fetched via same-origin /api/era5 (Open-Meteo customer archive; the API key is not in this page). Local evaluation may use Open-Meteo’s public archive. Weather data by Open-Meteo.com