Digital twin is used for many things in solar. On the SolarDetech platform it means one precise thing: a panel-level record. Every panel in the plant has a persistent identity, a geographic position, an electrical position and a measurement history, and the record is updated with every flight.
This article explains what that record holds, how the identity stays the same from one flight to the next, and what an asset manager, an O&M team or an investor does with it.
What the panel-level digital twin holds
- Persistent identity: the same identity for the same physical panel on every flight.
- Geographic position: block, row and table, plus the tier and column within the table.
- Electrical position: transformer, inverter, combiner and string.
- Measurement history: every flight's measurement of the panel, its findings with class and priority, and the raw radiometric frames behind them.
- Serial number: entered by scanning the label with a camera (OCR), and kept on the same position.
- Work orders: every job opened on the panel, with its status and notes.
The two positions answer different questions. The geographic address is the way to the panel on site; the field team finds a finding by it. The electrical address shows which string and which inverter a finding affects. A string fault, three or more consecutive modules running warm, only makes full sense when the record knows which modules belong to the same string.
How the identity stays persistent
The identity does not come from a sticker or a single image. Each flight is built into a georeferenced orthomosaic with RTK positioning, and panel segmentation records every module with its outline and its place in the plant layout. Because the identity is tied to that place (block, row, table, tier, column), it stays the same on every later flight.
A panel can change while its place does not. When a panel is replaced, the serial number of the new panel is entered on the same position. The replaced panel's serial number and its history stay on the record. Every panel also carries a state: Healthy, Anomaly present, In repair, Out of service or Decommissioned.
A record per panel, not a report per flight
A report per flight answers the question of what was wrong on that day. A record per panel answers the question of what has happened to this panel. Flights from different dates are compared on the same panel identities and at the same threshold, so new, growing and resolved anomalies show separately.
Three workflows depend on that comparison.
- After a storm: the flight after the storm is compared with the previous one. New physical damage and missing or detached modules are separated as new findings, with the panel each one is on.
- Before and after warranty: a flight before the warranty period ends puts a dated measurement of every panel on record. After a replacement or repair, the next flight measures the same panel again and the two records open side by side.
- Flight-to-flight time series: a panel's temperature history is followed from inspection to inspection, and growing anomalies stand out.
The comparison is fair because the conditions are recorded. The irradiance and weather of every flight are written into the report, so flights from different dates are compared on the same footing.
From finding to work order
A finding does not stay on the platform. It is listed with its class, priority, probable root cause, location and estimated production loss. A work order opens from it and is tied to the panel's record. The work order is assigned to a person on the plant team, the field team is guided from the map to the panel, and the status is tracked from opening to close-out. A fixed anomaly closes on the same panel record at the next inspection.
This is where the persistent identity pays off. Without it, the next inspection would produce a new list, and someone would have to match it against the old one by hand to see which repairs held.
Asset management on the same record
A solar asset management platform is only as good as the unit it manages. When the unit is the panel, the views above it can be exact.
- Portfolio: all plants on one screen, with the fleet map, each plant's last inspection date, its open findings and its priority breakdown.
- Production loss and efficiency: an estimated annual production loss for every finding, and plant efficiency by period and by anomaly group.
- Fire risk: anomalies that can carry fire risk go to the top of the list with very high priority.
- Warranty claims: the serial number, measurement history and raw frames are gathered into one claim file, and the verification flow with the manufacturer runs on the platform.
- Teams and roles: the field team, the operator and the investor reach the same panel record with separate roles.
- Integration: an API for partners to add their own users to plant teams and generate one-click sign-in links.
The full module list, with screens from the 8 MW demo site, is on the platform page.
What the twin is, and what it is not
The twin is a measured record. It changes when a panel is measured again, which happens with every flight. SCADA and inverter monitoring watch production totals at inverter or string level and stay in place; the panel record holds what a flight measured on each individual panel. The two are complementary.
The record is built from thermal and RGB imagery. I-V curve and insulation measurement are not part of it; they are offered optionally through a solution partner, and inverter faults are followed through SCADA.
How the record is fed
The engine behind the record is autonomous thermal inspection. Orthomosaic building, panel segmentation and anomaly detection run without human intervention, the model assigns each anomaly one of 17 classes and an expert confirms it, and every finding is verified across several raw radiometric frames before it is written onto a panel. The rules behind that, and the acceptance check every flight passes first, are on the methodology page.
Coverage matters here more than anywhere. A record per panel is only complete if every panel is measured, so the whole plant is flown and a flight that fails the coverage check is not processed. That is what 100% panel coverage means: the share of panels measured, not a detection accuracy figure.
The record can be fed by our team's flights or by your own RTK drone with a radiometric thermal camera. SolarDetech approves every inspection. Approval covers data acceptance and the upload link. Processing after approval is fully autonomous. The FAQ covers own-drone requirements, plants outside Türkiye and what is free.
To put your plant on the platform, send us your panel count and location. We reply within 24 hours with a quote. The platform itself is free; the inspection is priced by quote for your site.