In previous research, Artificial Intelligence (AI) and Machine Learning (ML) have been studied to enable and apply its ability to be utilized in the electricity grid. The overall result has been that AI has given a good result by contributing with better insights into the electricity grid. But much needs to be done to get to its full potential. Sweden is well on its way to being able to apply more AI in its own electricity grid. As new regulations have been introduced where the new generation of electricity smart meters will be able to collect a large amount of data (big data), which is the fuel for AI and ML.In this thesis, an machine learning algorithm is being experimented with to find electricity flow and consumption patterns in the low-voltage power system in Grönåker. With data from substations and electricity meters from the end customers as well as weather data. The project used programming and simulations via Python.The results that emerged were that the ML algorithm has the ability to find the right pattern of consumption in the electricity power grid in Grönåker. This pattern and insight can provide an opportunity to detect unreasonable meter values from end customers and major power losses that may be caused by, among other things, electrical thefts.