READING ON LOAD
Optimization of gas readings
Project to optimize the process of reading the gas supply points in order to reduce their operating costs, while maintaining the legal and quality requirements.

Detailed objectives:

Identification of supply points and cycles in which the reading of the route can be eliminated because the quality of the collection is not altered.

  • Redesign of the reading routes for theur optimization using the supplies identified for each cycle.
  • Development of a predictive model of consumption by supply and cycle that allows, in addition to identifying the supply points to be excluded from the routes, to enable new cases of future analytical use.

Data and volumetry:

  • Internal data: supplies and readings. Field data and special supplies. Volumetry > 1.5 billion records.
  • External data: Meteorological, sociodemographic and cartographic data.

Technologies:

Apache Spark, R / Shiny, Python, Scikit-kearn, Neo4j, Amazon Web Services (EC2, S3, Redshift, RDS), GIS.

Project type and dates:

  • Turnkey project.
  • Start date: Q3 2018.
  • End: Q1 2019.