INSUS learns theft and non-technical loss patterns from your consumption history and confirmed inspection outcomes, then gives revenue protection a prioritised list of accounts most likely to warrant a visit. People decide every inspection and account action.
01
Flags from consumption data you already bill against.
02
Missing reads and rare confirmed theft handled in the pipeline.
03
Advisory. People decide every inspection and account action.
The problem
Electricity theft and other non-technical losses are still often investigated through random or complaint-triggered field visits. Those visits do not systematically target the consumption anomalies theft produces: irregular or sustained drops, or use that behaves unlike similar customers. Confirmed theft is rare, reads can be missing, and inspection capacity is spread without a ranking, so hit rates stay low.
The shift
Inspectors still make the visit, and the utility still decides every action. What changes is which accounts they visit first.
Today
With INSUS
How it works
This fits OptiStudio as revenue-protection decision support beside existing metering, billing and inspection systems. Access is read-only. It never triggers disconnection or billing action.
Connect
Connect daily or periodic consumption from metering and billing, customer information for segment, tariff and meter type, and field inspection outcomes for confirmed theft and legitimate use.
Understand
Prepare consumption records, address missing readings and correct the imbalance between rare confirmed theft and legitimate use so the model is not simply learning that most customers do not steal.
Predict
Score theft or non-technical loss likelihood for each account from consumption patterns, with the pattern characteristics that drove each score.
Advise
Present a prioritised inspection list for the revenue protection team, with pattern detail to support case preparation. The team reviews every flagged account and decides whether to inspect.
Verify
Compare flags with actual field inspection findings and feed confirmed outcomes back so the model can be refined for this utility's customer base.
Getting started
For electricity distribution revenue protection teams. Delivered through OptiStudio as inspection-prioritisation decision support. Expert Services can help with data assessment and validation. Advisory only. No new meters from INSUS.
Related for utilities
Electricity theft detection sits beside transformer health index prediction and NIS2 cybersecurity compliance on the same network.
Theft and non-technical losses are often found by random or complaint-triggered inspection. This use case scores likely theft from consumption patterns and ranks accounts for field inspection review.
Historical daily or periodic consumption plus confirmed field inspection outcomes. No new meters from INSUS. Monthly-only reads need performance established on that coarser data.
No. The revenue protection team reviews every flagged account. Disconnection, billing adjustment and any other account action require human review and often a regulatory process.
No. External peer-reviewed results on public datasets are not INSUS results and do not transfer automatically to another utility's customer base.
Through OptiStudio as revenue-protection decision support. Expert Services can support data assessment, integration and validation with your revenue protection team.
By comparing flagged accounts with actual field inspection findings in the pilot segment. Benchmark precision does not translate directly into a field hit rate.
It needs consumption history and confirmed inspection outcomes. Accuracy is specific to each utility's customer base, meter infrastructure, read frequency and theft patterns.
Utilities · Revenue protection
Start with one segment and the inspection history behind it. The useful first answer is whether your own confirmed outcomes already leave a pattern the next visit list can use.
