AutoLeitura & AutoClara — Equatorial Energia
Equatorial Energia / ANEEL P&D
Highlights
- WhatsApp-native consumer interaction via virtual assistant Clara
- Fraud resistance through self-audit image capture
- Reduced meter-reader operational costs
- Real-time consumption tracking for consumers
The AutoLeitura project shifts the meter-reading burden from field operators to engaged consumers through a frictionless WhatsApp-based experience mediated by the virtual assistant Clara.
The consumer registers via email, receives a WhatsApp message prompting them to photograph their meter display, and Clara returns the recognized reading value for confirmation within seconds. The self-audit mode periodically requests photographic proof of the self-reported reading, using the same computer vision pipeline as SILEM to verify digit accuracy and detect physical tampering.
Expected improvements documented in the project: reduction in meter-reader headcount (leituristas), faster billing cycle, elimination of estimated-reading disputes, fraud minimization through the embedded audit capability, and seamless integration with pre-payment energy systems.