Short and straight — how Slunora arrives at a verdict on your system.
We take how much your system actually produced and compare it with how much it should produce given its location, tilt and orientation — using PVGIS, a public European model that accounts for real weather at that location. The gap between those two numbers is the basis of every verdict.
You either type in production by hand (copied from your inverter's app), or connect the inverter directly — we support SolaX right now, more brands are coming. A connected inverter is read automatically every day, so the verdict gets more precise over time: a rough estimate right after connecting, a monthly view after a few weeks, a full annual verdict after a year.
Every result comes with how confident we actually are — high, medium, or low. When something's missing (say, panel orientation), we say so plainly and lower the confidence — we never fill in gaps or guess on your behalf. Economic estimates are always shown as a range, never a single made-up number.
We're not a substitute for an on-site technical inspection — we can't physically check wiring, panels, or spot shade from a neighbour's tree. We tell you when an inspection is worth it, not that we've already done one for you.
We only read data, never change or configure anything on your inverter. Credentials are encrypted at rest, and you can disconnect and delete your data at any time.
Findings aren't one generic "problem" — they have categories and severity. We mainly watch for: long-term annual underperformance (production consistently below what the location offers), repeated weak months outside winter (a seasonal anomaly weather alone can't explain anymore), a month with practically no production (an outage — possibly a fault or lost communication with the inverter), suspiciously high output (usually a mis-entered installed power or orientation, not real overproduction), and specific yield outside the usual range for that location.
Diagnostics still work with incomplete data, just honestly. Without GPS coordinates we can't calculate expected output via PVGIS — the verdict stays indicative only. Without panel tilt and azimuth we assume optimal orientation, which overstates expected output for east-west or flat roofs — so confidence drops in that case. Without household consumption data we can't calculate self-sufficiency. Months with a data gap (the inverter didn't report) are excluded from annual totals but stay visible in the overview, so you know the gap is there.
Every finding carries two independent numbers, not one. Severity says how big the gap from expectation is — critical (a clear anomaly), needs attention (a mild gap worth watching), or informational (low-priority context). Confidence says how sure we are the finding is correct — high, medium or low, depending on how complete your inputs are. A critical finding with low confidence (say, without GPS) still shows up, but flagged clearly as worth verifying.