GridSentinel: Energy theft and loss analytics for power distribution
Find where energy goes missing.
GridSentinel turns smart meter data into an inspection-ready list of suspected theft and losses. It balances the energy received at every distribution transformer against what its consumers were billed, then ranks consumers by expected recovery so field teams go where it matters most.
Capabilities
What GridSentinel does
Transformer energy balance
Pinpoints the transformers and areas with unexplained loss.
Covers every loss type
Meter bypass, tampering, hooking, category misuse, overloading, billing irregularities and net-metering misreporting.
Fewer false alarms
Separates theft from faulty meters, communication gaps and vacant premises.
Explainable leads
Every flag comes with reasons and a recommended field action.
Closed loop
Inspection results feed back to improve the model over time.
Try it yourself
Balance a transformer the way GridSentinel does
Set the energy that reached a distribution transformer and what its consumers were billed. See the unexplained loss, its value, and where to send field teams first.
-
Ranked leads on this transformer
- Consumer #44817Night load, zero billed units for 3 cyclesLikely bypass: Inspect meter and incoming cable-
- Consumer #30922Billed as domestic, load profile matches a shopCategory misuse: Verify connection category on site-
- Consumer #51203Sudden 70% drop in consumption, meter seal eventPossible tampering: Check meter seal and terminal cover-
- Consumer #28844Readings missing for 9 days, then flatMeter or comms fault: Send meter team, not vigilance-
- Consumer #60117Solar export higher than panel capacity allowsNet-metering misreport: Verify installed capacity-
Illustrative sample. GridSentinel builds these leads from smart meter data and learns from your inspection results.
Value estimate
What could loss reduction be worth?
Enter your own figures. The recovery share is your assumption; a pilot measures what’s achievable on your network.
How it works
How it works
- 1
Balance
Energy in at each transformer is compared with energy billed to its consumers.
- 2
Locate
Transformers and areas with unexplained loss are ranked.
- 3
Explain
Suspect consumers are listed with reasons, separating theft from meter or data faults.
- 4
Inspect and learn
Field teams act on ranked leads; their findings retrain the model.
Built for
Who it’s for
- Electricity distribution companies (DISCOMs)
- Smart metering programmes
Getting started
What a pilot looks like
- 6–8 weeks on your own historical data, inside your environment
- Shadow mode: no impact on live systems
- Success criteria agreed upfront, with a clear go or no-go at the end
- Fixed fee, credited in full if you roll out
What data does GridSentinel need?
Smart meter readings, transformer-level energy data, consumer master data and billing records. Past inspection outcomes help the model learn faster.
Is it in production?
GridSentinel is pilot-ready and has been built and tested on simulated data. A pilot on your own data is the next step.
How does it avoid flagging honest consumers?
It separates likely theft from faulty meters, communication gaps and vacant premises, and every lead comes with reasons a field team can verify.
Where does the data go?
It stays in your environment or an approved cloud account.
More products
Also from Auto Solution
Stop fraud before the money moves.
Scores every payment in real time and flags likely fraud before it is approved.
For payment gateways, payment aggregators, fintechs
Explore PaySentinelEvery litre, accounted for.
Reconciles every fuel station four ways and flags each gap with its value and the action to take.
For oil marketing companies, fuel retail networks, fleet operators
Explore FuelLedger