Practical application of AI and Machine Learning

Reference number
Funding from Vinnova SEK 500 000
Project duration October 2019 - September 2020
Status Ongoing
Venture AI - competence, capacity and capability
Call Start your AI journey!

Purpose and goal

The project aims to increase the interest in AI / ML within Umeå Energy´s own organization, its customers and other stakeholders, and thus enable future business opportunities. By starting to use AI connected to the electricity consumption of private individuals, the project hopes to help them to reduce their energy use. The project goal is to discover and differentiate power-consuming units within a household based on the electricity meter´s collected measurement data at a high-resolution level (sec) using own hardware and at a low-resolution level (hour), the so-called free data.

Expected results and effects

In the short term, the solution will give the user the opportunity to influence his electricity usage by receiving analyzed and more clearly reported measurement data. In the longer term, this generates benefits for several parties, such as the electricity grids that receive lower load on the grid and can thus use existing infrastructure longer despite an increased need. The greatest potential exists if the project succeeds in the analysis of low-resolution measurement data, the so-called free data, which can contribute to a reduction in energy consumption for the entire community.

Planned approach and implementation

The project plans to use and further develop an already existing algorithm KNN (k-nearest neighbor) that has been used with good results in other studies. It is a relatively simple algorithm and the project can use the previous studies as reference material. The model will be trained in Microsoft Azure where the project has access to a large amount of high-resolution data through other already existing services, where the project will collect, anonymize and interpret measurement data from households. At least 2 households will act as reference points without anonymized data.

The project description has been provided by the project members themselves and the text has not been looked at by our editors.

Last updated 23 September 2019

Reference number 2019-03267

Page statistics