SIT4Energy — Smart Energy Intelligence
An intelligent energy management platform combining explainable AI, visual analytics, and behavioral science to empower prosumers in optimizing self-sufficiency, reducing consumption, and maximizing renewable energy utilization.
SIT4Energy develops intelligent, user-centered platforms that empower prosumers to optimize renewable energy self-consumption and utilities to forecast demand with explainable AI. The project integrates smart metering infrastructure, behavioral analytics, and adaptive incentivization to drive sustainable energy practices across residential and tertiary buildings.
Problem & Context
Traditional energy management focuses on supply-side solutions without considering prosumers who generate and consume their own renewable energy. Existing prosumer applications lack integration with local utilities, preventing optimization of grid demand. Consumers struggle with understanding consumption patterns, lack motivation for behavioral change, and utilities lack forecast tools tailored to small-scale operations.
Solution & Approach
SIT4Energy implements a comprehensive three-pillar strategy combining technology innovation, business model development, and consumer empowerment through explainable AI and visual analytics.
Prosumer Dashboard & Analytics
Real-time energy production and consumption visualization with self-sufficiency calculations and personalized recommendations. Interactive dashboards display hourly energy flows, cost savings, and comparative analysis enabling informed decision-making and behavioral change.
Utility Forecasting & Decision Support
Explainable k-NN algorithm for demand forecasting integrating weather parameters and historical patterns. Municipal utilities gain interpretable forecasts with 5% MAPE accuracy, supporting energy planning without requiring complex machine learning expertise.
Adaptive Incentivization & Behavioral Change
Personalized motivation strategies based on user value orientations (egoistic, altruistic, biospheric) and behavioral stages (pre-contemplation, contemplation, action). Context-aware notifications and recommendations trigger sustainable energy behaviors at decision-critical moments.
Mobile Services & Context-Aware Recommendations
Mobile app integration delivering consumption insights and AI-generated recommendations optimized for micro-moments of decision-making. Services adapt to weather conditions, device availability, and user engagement patterns for maximum effectiveness.
Pilot Implementation & Testing
SIT4Energy was validated in three pilot locations: residential prosumers in Haßfurt (Germany) with 22 participants and their solar installations, a smart home laboratory at CERTH/ITI (Greece), and tertiary building testing at Harokopio University (Athens). The 8-month open phase pilot (July 2020 - May 2021) provided real-world evidence of behavioral change and platform effectiveness across diverse user contexts.
Self-sufficiency improvement
Pilot locations (DE, GR)
Forecast accuracy (5.06% error)
Usefulness rating
Results & Impact
The SIT4Energy pilot validated the platform across three distinct settings: residential prosumer households in Hassfurt (Germany), the smart home laboratory at CERTH/ITI (Greece), and a tertiary building at Harokopio University (Athens). Over an 8-month open phase, the system demonstrated measurable impact on energy awareness and behaviour.
The explainable k-NN demand forecasting algorithm achieved 94.94% accuracy, providing municipal utilities with interpretable energy forecasts without requiring complex machine learning expertise. The adaptive incentivisation model — personalising recommendations based on user value orientations — and the context-aware reminder service contributed to a 5% improvement in energy self-sufficiency. 82% of pilot participants rated the platform as useful for managing their energy production and consumption.
EIPCM’s Role
EIPCM led the design and implementation of the Prosumer Dashboard with real-time energy visualizations, self-sufficiency calculations, and interactive analytics. The institute conceived a context-aware reminder service with adaptive incentivization that delivers personalized energy-saving recommendations at optimal moments based on behavioral insights and user patterns. EIPCM’s research focused on understanding how visual feedback, context-aware timing, and adaptive incentives most effectively motivate behavioral change while preserving user autonomy and trust.
Consortium
German Partners
- University of Applied Sciences Stralsund (Coordinator)
- Stadtwerk Haßfurt (Municipal Utility)
- EIPCM (Subcontracted Research Partner)
Greek Partners
- CERTH – Information Technologies Institute
- Information Technology for Market Leadership (ITML)
- Harokopio University (Tertiary Building Pilot)