SciTransfer
iFLEX · Project

AI Assistant That Automates Household Energy Use for Demand Response Programs

energyPilotedTRL 7

Imagine your home could automatically shift when it runs the dishwasher, heater, or charges the car — picking the cheapest, greenest moments — without you lifting a finger. That's what iFLEX built: a personal AI assistant that learns your habits, talks to your smart devices, and decides the best times to use energy on your behalf. It was tested with real families across three different climate zones in Europe, from Nordic cold to Mediterranean heat. The goal is to make participating in energy flexibility programs as effortless as setting your phone to "do not disturb."

By the numbers
13
consortium partners involved
3
climatic regions with field pilots
4
countries represented (DK, EL, FI, SI)
9
industry partners in consortium
42
total project deliverables completed
69%
industry participation ratio
The business problem

What needed solving

Energy utilities and retailers are losing money because residential customers won't participate in demand response programs — the manual effort is too high and the incentives too confusing. Meanwhile, grid operators desperately need household flexibility to balance growing renewable energy generation. Companies need a way to automate consumer participation in flexibility markets without creating friction or violating privacy regulations.

The solution

What was built

The project built a complete AI-powered software platform (the iFLEX Assistant) that autonomously manages household energy flexibility. Key deliverables include: a common software platform for creating customized energy assistants, natural user interfaces (iterated through 3 versions), an automated flexibility management module with AI optimization and rule-based safety controls, a hybrid deep-learning forecasting module for renewable generation and consumption, and a GDPR-compliant secure data management module — all validated through field pilots in 3 climatic regions.

Audience

Who needs this

Electricity retailers and aggregators running demand response programsSmart thermostat and home energy management system manufacturersDistribution system operators managing grid congestionProperty managers seeking to reduce residential energy costsEnergy service companies (ESCOs) offering flexibility-as-a-service
Business applications

Who can put this to work

Energy utilities and retailers
enterprise
Target: Electricity retailers or aggregators offering demand response programs

If you are an energy retailer struggling to sign up and retain residential customers in demand response programs — this project developed a ready-made AI assistant platform that automates flexibility management for households. Instead of asking customers to manually shift their energy use, the iFLEX Assistant handles it autonomously while respecting user comfort preferences. It was validated with field pilots across 3 climatic regions with real end-users.

Smart home and building technology
mid-size
Target: Smart thermostat or home energy management system manufacturers

If you are a smart home device maker looking to add energy optimization intelligence to your product line — this project built a common software platform for creating application-specific energy assistants with natural user interfaces. The system integrates deep learning forecasting, rule-based safety controls, and GDPR-compliant data management. With 9 industry partners already involved, the technology is designed for commercial integration.

Real estate and property management
any
Target: Property managers or housing associations with large residential portfolios

If you are a property manager looking to reduce energy costs across residential buildings — this project developed an automated flexibility management module that optimizes energy use across heating, cooling, and electricity without tenants needing to change their behavior. The AI learns consumption patterns and shifts loads to cheaper periods automatically. It was built with consumer co-creation to ensure tenant acceptance and comfort.

Frequently asked

Quick answers

What would it cost to deploy this in our customer base?

The project data does not include specific licensing or deployment costs. However, the iFLEX platform was designed as a common software platform for creating customized assistants, suggesting a scalable software-as-a-service model. Contact the coordinator at VTT (Finland) through SciTransfer for commercial terms.

Can this scale to thousands or millions of households?

The system was architectured as a common software platform specifically designed to create application-specific assistants for different services and markets. It supports customization for different incentive and market mechanisms across countries and climatic regions. Field pilots were conducted in 3 climatic regions, demonstrating cross-market adaptability.

Who owns the IP and can we license it?

The project was coordinated by VTT (Finnish national research centre) with 13 consortium partners across 4 countries. IP ownership is typically shared among consortium members under Horizon 2020 rules. Licensing arrangements would need to be negotiated with the relevant IP holders — SciTransfer can facilitate this introduction.

Does this comply with European energy and data regulations?

Yes — the project specifically built a GDPR-compliant secure consumer data management module providing trust management and data confidentiality services both at rest and in transit. The system was designed to work within existing European energy market structures and incentive mechanisms.

How mature is this technology — is it ready for commercial deployment?

This was an Innovation Action with field pilots completed in 3 climatic regions with real users. The consortium delivered final versions of the core platform, natural user interfaces, and automated flexibility management modules. The project closed in April 2024, meaning all 42 deliverables were completed.

Can this integrate with our existing smart home or energy management systems?

The iFLEX platform was specifically designed as a common software platform for building application-specific assistants. It includes a hybrid-modelling module for forecasting and modelling of flexible resources like heating and cooling systems. The natural user interfaces went through 3 iterations (initial, revised, final) to ensure usability.

Was this tested with real consumers or just in a lab?

Field pilots were conducted with real end-users in 3 different climatic regions across the consortium countries (Denmark, Greece, Finland, Slovenia). Consumer co-creation was built into different project phases, coordinated by a consumer organisation in the consortium. A/B testing of different incentive and engagement mechanisms was performed with real users.

Consortium

Who built it

The iFLEX consortium is heavily industry-driven with 9 out of 13 partners (69%) coming from industry, which is a strong signal of commercial intent. Coordinated by VTT, Finland's national research centre, the project spans 4 countries (Denmark, Greece, Finland, Slovenia) — covering Nordic, Central European, and Mediterranean markets and climates. The presence of only 1 university and 2 research organizations confirms this was execution-focused, not academic. With 2 SMEs in the mix and a consumer organisation handling end-user co-creation, the consortium was structured to move technology from lab to market. The geographic spread across 3 climatic regions means the solution was stress-tested in very different conditions — a critical selling point for any company looking to deploy across European markets.

How to reach the team

Coordinated by VTT Technical Research Centre of Finland — contact through SciTransfer for introductions to the project team and licensing discussions.

Next steps

Talk to the team behind this work.

Want to integrate iFLEX's demand response AI into your energy product or service? SciTransfer can connect you directly with the right consortium partner for your use case — whether that's licensing the platform, accessing pilot data, or exploring a commercial partnership.