Official call identifier: COSMIC
Objective of the call
The COSMIC 2nd Open Call supports SMEs, including start-ups, in developing, testing and validating data- and AI-based services and solutions for large-scale energy resource optimisation. The call aims to bring innovative third-party solutions into COSMIC’s AI-driven platforms and large-scale pilots, supporting the transition to a net-zero society through improved energy efficiency, resource optimisation, renewable energy integration and sustainable infrastructure management.
Scope of activities
The call funds the development, facilitation, testing and validation of innovative data/AI-based assets or services that directly support key energy resource optimisation applications. Each proposal must focus on one specific prioritised challenge from the Technical Guidelines. Proposals addressing multiple challenges in a single application will not be considered.
Eligible activities include solution development, integration with existing platforms, testing and validation in large-scale pilot environments, proof-of-concept and demonstration projects, and assessment of resources used in delivering the service or developing the solution. The proposed solutions may include AI and data-based services, tools and technologies such as digital twins, smart charging systems, user interfaces, BMS, BIM/GIS-based digital twins or Home Energy Management Systems.
COSMIC Open Call 2 challenges
• Challenge 20: Data-Driven Behaviour Modeling
This challenge focuses on data-powered consumer research to identify behavioural drivers, fairness perceptions and trust mechanisms that shape engagement with residential energy flexibility services. Projects should capture how different user segments perceive and respond to AI and data-driven services, using approaches such as surveys, discrete choice experiments, A/B testing, RCTs, AI or advanced statistical modelling. The results should support evidence-based service design, pricing and communication strategies for energy flexibility services.
• Challenge 21: Energy Literacy Online Game
This challenge aims to develop and launch a free, open-access web-based game that teaches energy flexibility, renewable energy production, system resilience and AI-enabled resource optimisation to a broad public audience. The game should help non-expert users understand renewable variability, demand-response, load shifting, small-scale production such as rooftop PV, and trade-offs between comfort, cost, emissions and grid stability.
• Challenge 22: Port Power Digital Twin
This challenge focuses on developing a real-time Digital Twin of the Port of Palma’s private electrical network. The solution should represent the current state of energy flows across the full MT/BT infrastructure, ingest data from heterogeneous sources such as meters, SCADA and IoT platforms, expose the network model to downstream systems, and provide a governed multi-user access layer for APB and concessionaires.
• Challenge 23: Port Energy Investment Simulator
This challenge aims to develop an energy investment scenario simulation tool for the Port of Palma. The tool should allow APB’s technical team to virtually define infrastructure modifications, such as new PV assets, BESS or grid upgrades, and evaluate their technical and economic impact against the current configuration using real historical data. It should also include backtesting to quantify what the proposed infrastructure would have achieved during past operating periods.
• Challenge 24: Port Energy Management System
This challenge focuses on developing a Port Energy Management System for the Port of Palma to optimise APB’s private electrical network in real time, intraday and day-ahead horizons. The system should minimise power demand from the DSO at each connection point, maximise collective self-consumption across the port, minimise APB’s overall energy cost, generate setpoints for controllable assets and calculate incentives linked to concessionaire generation and self-consumption.
• Challenge 25: Geo-AI-Driven Optimization of DHC Networks
This challenge aims to develop an automated workflow to generate and optimise district heating networks. The solution should minimise total network length while selecting buildings or communities to serve, respecting existing infrastructure and producing GeoJSON outputs with building footprints and optimal network layouts for simulation and GIS-based refinement. It should evaluate three district heating network scenarios per district against a minimum spanning tree baseline.
• Challenge 26: Uncertainty Quantification for DHC Networks modelling
This challenge focuses on developing a high-level ML/AI tool for uncertainty quantification in district heating and cooling networks. The tool should analyse how uncertainties in input data, such as building energy demand, climate conditions and energy prices, affect network layouts and KPIs including total cost of ownership, payback periods and energy efficiency. It should help identify which network sections need to be optimised under climate and market uncertainty.
• Challenge 27: Digital Twin of WWTP Biological Reactor
This challenge aims to develop a Digital Twin of a wastewater treatment plant biological reactor, combined with an AI-based multi-criteria optimiser. The solution should simulate different aeration strategies, predict oxygen demand, integrate renewable energy generation and day-ahead energy prices, and generate daily operation patterns that minimise energy costs and GHG emissions while maintaining water quality compliance.
• Challenge 28: Multi Vector Energy Optimiser in WWTP
This challenge focuses on developing an AI-based optimisation solution at wastewater treatment plant level. The solution should coordinate energy consumption, energy production, and import/export of energy to electricity and gas grids, considering electricity and heat consumption from plant processes and energy generation from solar PV and biogas. The aim is to support higher energy neutrality, lower energy costs and reduced environmental impact.
• Challenge 29: Energy Bill LLM Microservice
This challenge aims to develop an intelligent user-facing application or platform that enables consumers to submit energy bills as PDFs or photographs and receive AI-powered or automated analysis. The solution should explain bill components, detect anomalies or inconsistencies, benchmark consumption, provide personalised recommendations on tariffs or contracted power, and translate energy use into CO₂ and environmental impact indicators.
• Challenge 30: REC Discovery & Onboarding Platform
This challenge focuses on developing a user-friendly digital platform that simplifies discovery, simulation of joining benefits and expression of interest in Renewable Energy Communities in Portugal. Users should be able to enter an address or postal code and receive clear information about nearby active, planned or emerging energy communities, including eligibility, expected benefits and onboarding steps.
• Challenge 31: AI-driven Social Support Tool
This challenge aims to develop an integrated decision-support tool for AVRA planners and social workers. The tool should combine management, building, monitoring and social data to generate clear, prioritised recommendations for energy efficiency refurbishment and social interventions at building and household level, with a focus on reducing decision-making time, energy costs and improving thermal comfort for vulnerable households.
• Challenge 32: Maintenance Scheduling Optimizer
This challenge focuses on building a scheduling support tool that generates optimised maintenance plans based on predicted faults, operational constraints and energy production criteria. The tool should prioritise alerts, estimate the operational and economic impact of intervention scenarios, support decisions on preventive maintenance schedules and generate technician schedules considering logistics, resource availability and intervention priorities, while keeping operators in the decision loop.
• Challenge 33: AI - Accelerator for Urban CFD simulations
This challenge aims to develop, propose or integrate an AI-based add-on for OpenFOAM to accelerate computational time for urban flow simulations. The goal is to keep the complexity of the CFD model while using AI/ML to accelerate numerical convergence, making high-fidelity urban simulations more practical for early-stage design decisions related to thermal comfort, climate resilience and energy demand.
• Challenge 34: Long-Term Prediction Model for Energy Demand
This challenge focuses on developing a forecasting model that predicts heating, cooling, domestic hot water and electricity demand over a 20-year-plus lifecycle using limited early design-stage inputs. The solution should use regression or machine learning methods, historical energy consumption data and digital twin models to simulate hourly energy profiles and support long-term planning decisions.
• Challenge 35: Energy Community Marketplace
This challenge aims to organise a marketplace linking companies with excess energy from PV plants and companies willing to buy renewable energy. The solution should include an AI/ML-based tool for optimal association of consumers and producers into an energy community, plus a user-friendly GUI for consumers, producers and marketplace organisers to manage data access, preferences, proposed sharing conditions and energy sharing contracts.
Eligible applicants
Eligible applicants are legal entities established in EU Member States or Horizon Europe Associated Countries (described in General Annex A, General Annex B, General Annex C).
The call targets SMEs, including start-ups, bringing specific data/AI-based assets through the COSMIC Open Calls. The COSMIC consortium will select up to 27 SMEs across two Open Calls, with the 2nd Open Call aiming to select up to 13 beneficiaries.
The call is relevant for businesses developing data/AI-based services, tools and technologies for energy resource optimisation. The guide specifically refers to environmental technology innovators, waste management and recycling tech SMEs, and user insight and behavioural analytics SMEs. COSMIC partners, their affiliated entities, and their employees or associates, whether current or former if previously involved in project execution, are not eligible to apply.
Eligible costs
The call applies a lump sum funding model.
Eligible cost categories include:
• Personnel costs
• Travel and subsistence
• Equipment (depreciation)
• Other goods and services
• Indirect costs
Funding conditions
Type of action:
Cascade Funding
Funding rate:
• Up to 100% for non-profit legal entities
• Up to 70% for profit-making legal entities
EU contribution per project:
Up to €150,000
Project duration:
10 Months
Technology Readiness Level (TRL):
Proposals must aim to reach TRL 6–7 by the end of the project. Solutions should start at a minimum of TRL 3, with initial validation at small scale, typically TRL 3–5, and be ready for further testing, refinement and demonstration through COSMIC’s industrial pilots.
*Technology Readiness Levels (TRL) describe the maturity of a technology, from basic research (TRL 1) to fully deployed, market-ready solutions (TRL 9).
Deadline for submission
20 July 2026