If you are a DR aggregator struggling to coordinate thousands of distributed energy assets manually — this project developed an AI-driven management platform that automates load profiling, forecasting, and dispatch using deep reinforcement learning. It was piloted across residential and commercial sites covering over 11 GWh of loads and over 14 GWh of renewable generation. The system uses open protocols like OpenADR for plug-and-play integration with existing infrastructure.
Smart Demand Response Platform That Lets Energy Aggregators Scale Efficiently
Imagine you run an energy company and need thousands of homes and businesses to use less electricity during peak hours — but coordinating all of them is like herding cats. DELTA built an AI-powered platform that automatically figures out who can cut back, by how much, and when, without anyone having to lift a finger. It uses machine learning to predict energy patterns and blockchain to keep all transactions secure. Two real-world pilots in the UK and Cyprus tested it across buildings, solar panels, and batteries.
What needed solving
Energy aggregators today struggle to manage thousands of distributed energy resources — rooftop solar, batteries, smart appliances — using centralized systems that become slow and expensive as they scale. Meanwhile, building owners and prosumers find demand response programs too complicated to participate in, leaving gigawatts of flexibility untapped. The result is missed revenue for aggregators and unnecessary grid stress during peak demand.
What was built
DELTA delivered an integrated prototype of an AI-powered demand response management platform with 39 deliverables. Key components include deep reinforcement learning for flexibility profiling, multi-agent energy matchmaking algorithms, multi-factor forecasting, blockchain-secured smart contracts for prosumer transactions, and a social collaboration platform — all tested across two pilot sites in the UK and Cyprus.
Who needs this
Who can put this to work
If you are a utility dealing with grid instability from growing renewable penetration — this project built multi-agent matchmaking algorithms that balance supply and demand clusters in real time. The pilots in the UK and Cyprus included energy storage systems of over 9 MWh, proving the platform can handle both implicit and explicit demand response services. Blockchain-based smart contracts secure all aggregator-to-prosumer transactions automatically.
If you are a facility manager looking to monetize your buildings' energy flexibility without disrupting tenants — this project created a fully autonomous system that manages demand response participation behind the scenes. Building occupants experience no disruption while the system automatically adjusts loads. The platform was tested with real residential and tertiary buildings across two pilot sites in the UK and Cyprus.
Quick answers
What would it cost to deploy this platform?
The project received EUR 3,873,625 in EU funding across 11 partners over 3.5 years. Deployment costs for a commercial version would depend on scale, but the platform was designed for scalability and computational efficiency — specifically to reduce the cost of managing large numbers of distributed assets. Contact the consortium for licensing or partnership pricing.
Can this handle industrial-scale operations?
Yes. The pilots in the UK and Cyprus covered over 11 GWh of residential and tertiary loads, over 14 GWh of renewable generation, and over 9 MWh of energy storage. The architecture was specifically built to distribute intelligence across multiple layers, making it scalable beyond what traditional centralized aggregator systems can handle.
What is the IP situation and how can I license it?
The project produced 39 deliverables including an integrated prototype. IP is held by the consortium of 11 partners. The coordinator CERTH (Greece) would be the first point of contact for licensing discussions. With 5 industry partners including 5 SMEs, the consortium was structured with commercialization in mind.
Does this comply with current energy market regulations?
DELTA was designed to work within current market rules while also pushing regulatory limitations around flexibility activation. It uses OpenADR, a well-known open protocol, for interoperability. The blockchain and smart contract layer was built specifically to meet data security and transaction authentication requirements.
How long would integration take with our existing systems?
The platform was built end-to-end using open protocols like OpenADR specifically to increase interoperability with existing energy management systems. The modular architecture means individual components (forecasting, profiling, matchmaking) can be adopted independently. Based on available project data, the integrated prototype was delivered and tested in real pilot environments.
What makes this different from other demand response solutions?
DELTA distributes AI intelligence to the edge rather than keeping everything centralized, which reduces computation bottlenecks as you scale. It combines deep reinforcement learning for flexibility profiling with multi-agent matchmaking algorithms — not just simple on/off scheduling. The blockchain layer for secure transactions is an additional differentiator over conventional platforms.
Is the project still active and who maintains it?
The project closed in October 2021. The consortium of 11 partners across 8 countries may still maintain components or offer commercial versions. CERTH, the Greek research center that coordinated the project, is the recommended first contact for current status and availability.
Who built it
The DELTA consortium brings together 11 partners from 8 countries (Austria, Belgium, Cyprus, Greece, Spain, Ireland, Norway, UK), with a healthy 45% industry ratio and 5 SMEs involved. This geographic spread covers key European energy markets with different regulatory environments, which strengthens the platform's adaptability claims. The mix of 5 industry partners, 3 universities, and 2 research organizations — coordinated by CERTH, a major Greek research center — suggests a team that balances academic rigor with commercial intent. The strong SME presence indicates the technology was developed with market deployment in mind, not just academic publication.
- ETHNIKO KENTRO EREVNAS KAI TECHNOLOGIKIS ANAPTYXISCoordinator · EL
- HIT HYPERTECH INNOVATIONS LTDparticipant · CY
- ARCHI ILEKTRISMOU KYPROUparticipant · CY
- C.C.I.C.C. LIMITEDparticipant · IE
- KIWI POWER LTDparticipant · UK
- E7 GMBHparticipant · AT
- UNIVERSITY OF CYPRUSparticipant · CY
- NORGES TEKNISK-NATURVITENSKAPELIGE UNIVERSITET NTNUparticipant · NO
- UNIVERSIDAD POLITECNICA DE MADRIDparticipant · ES
- CARR COMMUNICATIONS LIMITEDparticipant · IE
- JRC -JOINT RESEARCH CENTRE- EUROPEAN COMMISSIONparticipant · BE
CERTH (Ethniko Kentro Erevnas Kai Technologikis Anaptyxis) in Greece coordinated the project. Search for DELTA H2020 coordinator contact to find the project lead.
Talk to the team behind this work.
Want an introduction to the DELTA team? SciTransfer can connect you with the right consortium partner for your specific use case — whether you need the AI engine, the blockchain layer, or the full platform.