A company in the distribution sector is faced with the challenge of implementing an analytical model for the analysis, management and prediction of its consumption. (meters, analysers, probes, , etc.), in order to identify problems early and efficiently reduce its energy costs.
Objectives
Analysis of consumption by plant/zone at quarter-hour intervals, comparing forecasts, historical data, billing data and data from plants with similar characteristics. Analysis of the impact on consumption due to variations in average temperatures by geographic area. Homogeneous comparison of consumption behaviour among the plants in its network, in relation to the characteristics of its area, sales figures and other external factors.
Use case
Analysis, management and prediction of their consumption (meters, analysers, probes, , etc.), in order to identify problems early and efficiently reduce their energy costs.
Infrastructure
On Premise
Technology
Automatic or deep learning
Data
Public and private datasets
Resources
Multidisciplinary team with experts in machine learning, data analysts and data scientists.
Difficulties and learning
Difficulty in locating quality public data with which to contextualise their influence on the company’s consumption data.
KPIs (business impact and metrics of the model)
Improvement in energy consumption per sales and sales area of each plant