Energy & Utilities

A leading energy company

Data Science for renewable energy


In the context of a Data Science consulting assignment, Astrakhan is working with a leading energy company on projects related to renewable energy. So far, our work focused on photovoltaic solar energy projects and electric vehicles, but we have also worked on purely technical projects such as statistics and computer development.

Our achievements

● Implementation of automated and daily calculations of metrics useful for the study of different photovoltaic panels
● Modeling and forecasting the power generated by a photovoltaic panel with weather data
● Contribution to the team’s Git to share and group code files from different projects
● Involvement in the team’s technology watch: sharing new Data Science libraries, frameworks…
● Implementation of a proof of concept on a sample of data using methods for sensitivity analysis of parameters in a complex physical system
● Implementation of anomaly detection algorithms on time series of power generated by solar energy
● Creation of an algorithm to recognize certain types of roofs from satellite images to identify buildings that can accommodate photovoltaic panels
● Implementation of algorithms for short-term forecasting (D+1/D+3) of power used by charging stations for electric vehicles

Client Benefits

Our work allows our client to develop its expertise in renewable energy, and to benefit from the following:
● Development of a roof type recognition tool to help customers determine the most interesting buildings in terms of photovoltaic panel installation
● Access to daily physical metrics to facilitate the work of business experts
● Improved predictive models for power generated by photovoltaic panels
● Development of algorithms to better predict electricity consumption over the day, in particular at charging stations for electric vehicles
● Code files allowing a first approach to different useful methods in Data Science, such as cleaning time series data or anomaly detection
● Contribution and consolidation of theoretical and technological knowledge in Data Science and Statistics

Our added value

●Integration of quality scripts adapted to the existing client infrastructure
●Automation of analyses, transformations, data cleaning
●Skills on the entire data chain, in Data Science as well as in Data Engineering
●Clear and accurate presentations of the results obtained