This subsection is a short energy data analysis and explores the energy data coming from smart meters. The smart meter data are usually massive, and a big data infrastructure is valuable and support to analyze them. Furthermore, we have to aggregate the smart meter data with other data sets like weather data or socio demographic characteristics. Therefore, the actions and examination in this sub-sectionsubsection could be performed in a distributed system. Energy systems have important elements like generation, transmission, and distribution. Smart grids are at the distribution level, and generation and transmission are at earlier stages. The data we consider is household, and Figure 12 presentpresents these elements.
The data collected at these systems are limited in availability. Similarly, the amount of data could be enormous. There are challenges atin these IoT systems, and we can ask how we can make data analysis in IoT systems when faced with these difficulties. An essential device in multiple smart energy systems is a smart meter that produces data. ThisThese data isare the energy consumption in a time dimension and are raw and immature data. However, the real-time information is the data from smart meters but cleaned and stored. Moreover, these data have accuracy and high availability [134]. Both the RCTs and the Smartsmart meter data do not have labels. The RCTs are explained atin the deliverable 1.1 Analysis of best practices atin section 4.1.1. The only information that provides us is the energy consumption in the time dimension, after we can aggregate additional information according to the properties of the consumer. Furthermore, we have to apply ML methods to classify the data.