Goals

GREEN.DAT.AI aims to develop novel Energy-Efficient Large-Scale Data Analytics Services, ready to use in industrial AI-based systems. The project will demonstrate the efficiencies of the new analytics services in four industries: Smart Energy, Smart Agriculture/Agri-food, Smart Mobility, Smart Banking and six different application scenarios, leveraging the use of European Data Spaces.

The ambition is to exploit mature solutions already developed in recent H2020 projects and deliver an efficient, massively distributed, open-source, green, AI/FL ready platform, and a validated go-to-market Toolbox for AI-ready Data Spaces.

The GREEN.DAT.AI Toolbox will be by design compliant with the FAIR data and metadata management principles. In the long-term, the GREEN.DAT.AI platform will allow computing to move from data centers to edge devices, making AI accessible to more people, shift computation from the cloud to personal devices to reduce the flow and potential leakage of sensitive data and enable processing data on the edge to eliminate transmission costs, leading to faster inference with a shorter reaction time and drive innovation in applications where these parameters are critical.

The GREEN.DAT.AI Consortium consists of a multidisciplinary group, well balanced in terms of expertise. GREEN.DATA.AI applies a multidisciplinary approach to draw on knowledge from different domain experts in energy, transport business and economics as well as Data Science and SW Engineering.

Highlights

• Project coordination by INLECOM Innovation
• Consortium formed by 17 partners from 10 different countries (and one associated party)
• The six demonstration solutions will occur two in Portugal (wind farms and electric vehicle charging data), two in Spain (electric mobility and banking cybersecurity data) and two in Slovenia (smart farming optimization and agrifood water management).

Demonstrators

The objectives of the demonstrators are:

1. provide an experimentation and innovation environment and testbed for GREEN.DAT.AI solutions in six pilots,
2. define the pilot’s implementation plan by further detailing the User Stories and specifications,
3. implement co-creation actions and co-delivery plan and monitor progress, (4) provide outputs to WP6 for exploitation and replicability.

The Green.Dat.AI solutions will demonstrate the efficiencies of the new analytics services in six pilots from four industries: Smart Energy, Smart Agriculture/Agri-food, Smart Mobility, Smart Banking:

1. Energy Marketplaces: Data sharing across the renewable energy sector – This demonstrator will build on previous work from H2020 Smart4RES project to develop algorithmic solutions for data markets, which will allow different agents to sell and buy data (but also considering the possibility of data barter trading) of relevance for RES forecasting while being reattributed for it.
2. Energy Grids: Smart electric vehicle charging – It will be built on previous work: (a) Data cleaning and enrichment pipeline, (b) Unified NoSQL data access operators for NoSQL stores and (c) trajectory offline/online analytics suite; to increase maturity until TRL 7-8, by redesigning algorithms to efficiently perform at the fog/edge (de-centralised analysis of routes, distances, usage, average times of use, etc.), while CNR will work on predicting individual EV energy demand in space (in which station) and time (when to recharge) based on recent GPS traces of vehicle and INESC TEC will work on ML and AI to support predictive edge-based management and control of clusters of locally positioned EV in a new generation of EVSE systems.
3. Agriculture: Smart Farming Optimisation through Digital Twins – It will adapt UAV video-stream (pre-) processing techniques to be carried out at mobile devices controlling it (UM). UM will also develop knowledge integration methods that allow for combining models learned to recognise diverse soil health statuses and detection of different pest at different geospatial locations and by different users.
4. Agri-food: Smart Water Management – It will use sensors for measuring water elements at various water sources and at different locations. The new services will improve existing methods for water quality monitoring, by using real-time IoT data from collection/monitoring sensors and processing/combining them with open water data from national databases.
5. Smart Mobility: Energy Demand Response in EDVs & Infrastructure Maintenance – It will optimise the energy demand of the network on the bicycle stations using novel ML/AI prediction techniques. Energy demand forecasts will use historical data for patterns and anomalies identification, which will improve charging scheduling efficiency.
6. Banking: Fraud detection and explainable AI - Real-time or near real-time detection of known fraud signatures will be achieved via this demontrator. TSI’s accelerated pattern matching engine will be adapted to provide real-time or near real-time detection of known fraud signatures, aiming the reduction of the stress on the AI/ML/Feature Learning as it can filter out common frauds that can be detected without engaging the computationally and energy expensive AI algorithms.

Role EDP NEW

EDP NEW R&D will lead the pilot Energy Marketplaces: Data sharing across the renewable energy sector and participate in the smart EV charging pilot.

EDP NEW R&D will supply data and infrastructures to promote these two pilots and design the replication and exploitation plan leverage by the business goals of EDP Group.

Partners