Skip to content
EtaTune recommendation screen with forecasted power generation chart

S-02 · EtaVolt · 2024

EtaTune

ML health prediction for photovoltaic systems, with remediation advice and earnings forecasts.

Discipline
Software
Year
2024
Company
EtaVolt
Role
Full-stack engineer
ReactChart.jsFastAPIpandasNumPyMatplotlibpvlib

EtaTune uses machine learning to analyse and predict the health of photovoltaic systems. By ingesting historical inverter data, it identifies degradation patterns and pinpoints their root causes.

From those insights, EtaTune recommends targeted remediation such as regeneration or panel replacement. It forecasts future generation both with and without remediation so users can compare outcomes, and calculates the financial impact of each option to support data-driven decisions.

Key contributions

  • Better prediction accuracy by factoring in historical weather and irradiance data
  • Complete refactor of the API server
  • New fire-safety score calculation
  • Unit testing

All rights reserved EtaVolt Pte Ltd.