operating_verdicts
Is that factory actually operating? We test operational status independently, using night-time-lights and thermal signatures, always validated against a real confirmed-operating site and a real confirmed-not-operating site first.
Real sample
Real B-05/B-06 method-validation battery, 2026-09-02 - optical/SAR (Sentinel-2/Sentinel-1) vs. thermal (Landsat), tested at the method-validation site (Tesla Gigafactory Grünheide, confirmed operating) and the negative control (Intel Magdeburg, project cancelled, never broke ground):
| Site | NDVI shift (σ) | NDBI shift (σ) | Backscatter shift (σ) | Thermal shift (σ) | Verdict |
|---|---|---|---|---|---|
| Grünheide (confirmed operating) | -2.99 | +2.95 | +1.89 | +1.01 | Optical/SAR: OPERATING_SIGNAL_DETECTED |
| Magdeburg (negative control, cancelled) | -0.62 | +0.21 | -0.01 | +0.15 | Correctly: NO_SIGNAL |
Optical/SAR method validated (strong real signal at the confirmed-operating site, clean negative control). The thermal channel was tested too and came back a clean, pre-registered FAIL (threshold was 2.0σ, Grünheide only reached 1.01σ) - reported honestly as a dead line, not hidden. Full detail: china_industrial_reality product page.
Method
Sentinel-2 NDVI/NDBI + Sentinel-1 VV backscatter time series, sustained shift in the last third of the post-baseline period measured in units of pre-registered baseline sigma. Threshold fixed before looking at any result.
Request access / notify me
This line has real, active work behind it - not a placeholder. Leave your email and we'll reach out when it graduates to a product.