The SIT4Energy project is demonstrating the idea that a smart IT solution can be used in the implementation of integrated energy management for optimizing energy production and consumption patterns. SIT4Energy considers efficiency potentials in the local energy production and consumption. Subsequently, several components are under development offering solutions in different fields.

SIT4Energy aims to support behavioral change towards adopting a more environmentally friendly attitude. However, as static recommendations have seemed inefficient to grab user attention, a novel recommendation system needs to be implemented. In brief, the following main subjects and their corresponding solutions are proposed, which are on an ongoing state and under continuous progress.

  • SIT4Energy Recommendation Engine: Stimulate actual empowerment of consumers and increased understanding of sustainable energy and energy billing
    The SIT4Energy Recommendation Engine (RE) is going to be used as core for SIT4Energy Recommendation making processes. A recommender system can be defined as any system that guides a user in a personalized way to interesting or useful objects in a large space of possible options. In the context of smart homes, recommendation technologies improve the overall applicability of the installed equipment and can also help to optimize the usage of the available resources (e.g., minimizing power consumption), without decreasing the comfort level of the inhabitants. The SIT4Energy Recommendation Engine prototype aims to deliver accurate personalized recommendations based on the combination of the type of energy consumption together with additional environmental data derived from the installed smart sensors. It is established on an initial rule-based system and it is extended by incorporated machine learning techniques originated by additional user profiling and comparison of historical consumption data. Its predefined conditions trigger a recommendation to be sent to the mobile app when these prerequisite conditions are met. At the current stage, initial developments are made based on the data retrieved from the SIT4Energy academic pilot, where several smart meters and sensors are already in place and further upgrading activities are currently taking place.
  • Implement context-based mobile recommendation services for energy end-users based on user micro-moments and other means of feedback to consumers
    The initial requirement for the beginning of the SIT4Energy mobile Recommendation Engine was the identification of the most applicable micro-moments for pushing energy related recommendations and incentives. In the context of the project, the term micro-moment is introduced, to distinguish those moments in which the user would be most receptive to a recommendation regarding the energy management of his building. Extending the existing definition, a micro-moment could be redefined as the moment in which either the user’s activity is being still or tilting his phone or the phone screen is being turned on while being in his house. The activity detection and the subsequent micro-moment exploitation has been achieved through developing a novel Deep Neural Network model, which combines a simple feature extraction and Convolutional Layers, able to recognize human physical activity in real time from tri-axial accelerometer data. The goal is to run inference in real time on a mobile device.
    The main functionalities of the mobile application include providing the users with an option of a consumption screen offering the graphical representation of their consumption per selected timeframe as well as an analysis of their energy consumption values per category (i.e. light, HVAC or other equipment). Moreover, the application includes a condensed view related to user energy savings and predefined Kwh Target consumption. This functionality displays the monthly savings (in euro) based on monthly energy consumption limit that has been defined by the end-user, considering the electricity rates which will be initially predefined to a constant value. Both Recommendation Engine and Mobile Application are continuously enriched towards the delivery of the holistic SIT4Energy Recommendation Engine and Mobile Application included with “adaptive incentive service”, “context- aware attention triggering service” as well as “activity tracking and micro-moments services” developed by SIT4Energy consortium.
  • Smart Energy Dashboard
    The main objective is to aid utility analysts in better managing of local energy demand and supply to support the decision-making process. The smart energy dashboard combines explainable machine learning methods such as kNN with interactive visualization and explorative analysis. In this way, utility analysts can understand how different factors influence expected consumption to assess possible demand forecast under certain conditions.
    The prototype is designed to use the data from prosumer’s smart meters to provide interactive visualizations of their energy consumption and production as well as to provide recommendations on how to increase their self-sufficiency. Future versions of the dashboard will incorporate weather data and energy prices.