RETRO-RO - Part B Draft
HORIZON-CL6-2026-02-CLIMATE-02: Towards the water infrastructures of the future (RIA, single-stage)
Executive Summary
Abbreviations: reverse osmosis (RO); per- and polyfluoroalkyl substances (PFAS); antimicrobial resistance (AMR); nature-based solutions (NbS); artificial intelligence (AI); Earth observation (EO); information and communication technologies (ICT); operational technology (OT); Water Framework Directive (WFD); Technology Readiness Level (TRL); key performance indicator (KPI); work package (WP); greenhouse gas (GHG); FAIR (Findable, Accessible, Interoperable and Re-usable); INSPIRE (Infrastructure for Spatial Information in Europe); Water Security for the Planet (Water4All); information and communication technologies for water (ICT4Water); General Data Protection Regulation (GDPR).
RETRO-RO develops and validates a retrofit resilience toolkit that enables drinking water and wastewater infrastructures to remain safe, low-carbon and secure under climate-driven extremes, emerging contaminant pressures and malicious threats. The project targets the installed base of critical water assets and delivers modular, procurement-ready 'upgrade packages' rather than new-build infrastructure. It explicitly addresses compound risks from climate-driven hydrological extremes and from increased human mobility, which can accelerate the introduction and spread of waterborne pathogens and antimicrobial resistance in source waters.
Module 1 - Retrofit Resilience Layer for legacy treatment trains (demonstrated on reverse osmosis (RO) units): paired monitoring across source-water intake, product water and reject or concentrate streams; operator-approved set-point recommendations constrained within predefined safe operating envelopes; auditable decision support compatible with utility supervisory systems; compliance-first management of reject streams with conditional circularity when safety gates are satisfied.
Module 2 - Nature-based Solutions (NbS) as Hydraulic & Pollution Buffers: design and integration of NbS to reduce peak inflows and pollutant pulses entering sewers and plants, reduce carbon footprint and support biodiversity.
Module 3 - Digital Water Intelligence: secure digital twin and artificial intelligence (AI) predictive analytics integrating in-situ sensors, smart meters and Copernicus-derived Earth observation indicators with dynamic catchment-to-intake source-water risk mapping for chemical pollution and microbial hazards (bacteria, antimicrobial resistance and viruses). The mapping covers both groundwater-fed sources (springs and wellfields, including recharge-zone delineation and vulnerability assessment) and surface-water abstractions (rivers, lakes and reservoirs, including event-driven transport and resuspension processes). Outputs are confidence-scored hotspot and pathway maps and early-warning alerts that trigger auditable operational playbooks (e.g., intake switching/blending, targeted sampling escalation, and bounded treatment adjustments on existing reverse osmosis units and related assets). The digital twin is designed for interoperability with Copernicus services and with European digital-twin initiatives on freshwater and the Earth system, including the European Union Digital Twin on freshwater and Destination Earth.
Module 4 - Resilience-by-Design Security & Data Collaboration: OT cybersecurity baseline, physical-cyber continuity procedures, and a secure INSPIRE-aligned data-sharing framework mapping interdependencies with energy and ICT infrastructures.
RETRO-RO validates the toolkit through a pre-registered stress-test ladder (TRL2-3 to TRL5) and delivers falsifiable KPIs against matched baselines. Key exploitable results are packaged as commissioning and acceptance-test scripts, interface specifications, SOPs, cybersecurity baselines, data models and FAIR datasets, plus a good-practice guideline for new or renewed infrastructures (resilience, redundancy, ecological and social sustainability).
1. Excellence
1.1 Objectives
Derive and calibrate observable indicators, thresholds and operating envelopes that classify stress regimes (hydraulic variability, flood/drought triggers, emerging contaminants (including per- and polyfluoroalkyl substances (PFAS) and microplastics) and biological risk (including antimicrobial resistance and waterborne pathogens such as bacteria and viruses)) and provide actionable early warning.
Design and implement a modular retrofit architecture spanning treatment assets (demonstrated on legacy reverse osmosis units and advanced wastewater treatment assets), upstream nature-based solutions buffering and network-relevant monitoring (leakage, infiltration and inflow), enabling bounded adaptation without disrupting certified core equipment.
Develop a secure digital twin and AI layer integrating in-situ sensing, smart metering and Earth observation indicators to support predictive operation, risk-based maintenance, energy/GHG minimisation and evidence-based decision-making.
Develop dynamic source-water pollution and microbial-hazard maps for catchments feeding critical intakes (including springs, wells and surface-water intakes in rivers/lakes/reservoirs), combining Earth observation, historical WFD/utility records and event-triggered sampling to identify hotspots, transport pathways and early-warning triggers under flood/drought regimes. The mapping explicitly accounts for climate-driven mobilisation and for upstream pressures linked to land-use change and human mobility (tourism and travel).
Demonstrate resilience, safety and security through a pre-registered stress-test ladder (lab to pilot), including cyber-physical disturbance drills; report falsifiable KPIs against matched baselines.
Deliver replication packages and good-practice guidelines: procurement-ready specifications, acceptance tests, SOPs, cybersecurity baseline, INSPIRE/FAIR-aligned data model, and stakeholder adoption pathways.
1.2 Concept and overall approach
RETRO-RO is designed around the topic's core requirement: infrastructures must be flexible under flow and pollution variability, integrate digital solutions and citizen science, be protected against cyber/terrorist threats, follow good practices for resilience and sustainability, and improve design through better prediction and mitigation of floods and droughts. The project therefore operates at two coupled scales: (i) asset-level retrofit demonstrators (treatment and monitoring modules) and (ii) system-level governance and design guidance (data sharing, interdependency mapping, emergency operation and replication).
The strategic choice is to prioritise 'deployability': modules are engineered as retrofit add-ons with clear interfaces, acceptance tests and commissioning scripts so that utilities can procure and deploy them on legacy assets. Innovation is expressed as verifiable behaviour under stress (bounded adaptation with auditable evidence) rather than as a claim of novel plant construction.
1.3 Compliance matrix with expected outcomes and scope
Note: WPs referenced are defined in Section 3 (Implementation).
1.4 Editorial methodology and key innovations
1.4.1 Modular retrofit architecture
RETRO-RO implements four coupled retrofit modules, designed for minimal disruption to existing assets and for utility-grade commissioning:
Sensing and observability (paired upstream-to-downstream monitoring): abstraction characterisation plus reject-stream signals for early warning; multi-parameter sensing with periodic laboratory anchoring.
Envelope-bounded actuation: bounded pressure/recovery/crossflow and pre-defined recovery procedures (flush/relaxation) within pre-registered limits; operator-supervised actions only.
Auditable decision layer: supervised decision support that records recommendations, overrides, data provenance and protocol deviations in a tamper-evident audit log.
Compliance-first concentrate routing: interface and decision rules supporting safe downstream handling; conditional circularity routes only when effect-based toxicology and microbiological gates (including viral screening where relevant) are satisfied.
1.4.2 Nature-based Solutions integration
NbS components are selected and designed as 'upstream buffers' that reduce peak inflows and pollutant pulses entering sewers and plants, lowering stress on treatment processes and reducing carbon footprint. RETRO-RO will quantify NbS co-benefits (hydraulic peak attenuation, pollutant retention, biodiversity indicators, carbon footprint) and define practical integration patterns with grey infrastructure.
1.4.3 Digital twin, AI and Earth observation
RETRO-RO develops a secure digital twin that fuses in-situ monitoring (treatment, networks and nature-based solutions), smart metering and Copernicus-derived Earth observation indicators to deliver predictive analytics under flood and drought regimes. The twin is explicitly operational: it generates scenario triggers, recommended bounded actions, maintenance planning signals and evidence packages suitable for utility supervision and audit.
A dedicated catchment-to-intake source-water risk intelligence layer produces dynamic pollution maps upstream of critical abstraction points and links them directly to infrastructure operation. It is implemented as two coupled modelling streams that are harmonised into a common risk representation (hotspots, pathways, confidence-scored alerts and uncertainty bounds):
Groundwater-fed sources (springs and wellfields): delineation of recharge zones, vulnerability assessment, travel-time and attenuation reasoning, and event-driven mobilisation risk under drought and recharge pulses; integration of hydrogeological context with monitoring records and event-triggered sampling.
Surface-water abstractions (rivers, lakes and reservoirs): event-driven transport and dilution modelling, sediment resuspension and first-flush effects during floods, and concentration effects during drought and heat; assimilation of hydrological indicators and upstream monitoring into dynamic pathway maps.
Operational coupling is mandatory: mapping outputs are translated into auditable triggers and playbooks that guide intake and treatment decisions on the installed asset base. Examples include intake switching or blending, targeted sampling escalation for confirmation of emerging contaminants, activation of nature-based buffering measures where available, and bounded adjustments to pre-treatment and reverse osmosis operation (within predefined safety envelopes) to protect performance and water safety.
All mapping outputs carry explicit confidence scoring, data provenance and uncertainty reporting. Non-sensitive in-situ data products follow Infrastructure for Spatial Information in Europe (INSPIRE) principles and FAIR (Findable, Accessible, Interoperable and Re-usable) practices; security-sensitive operational data are shared under controlled access in line with critical-infrastructure protection requirements.
1.4.4 Citizen science and consumption monitoring for pricing
Citizen science is implemented as a controlled, quality-assured channel for early anomaly reporting and community-supported monitoring (e.g., event reporting, optional guided sampling where appropriate). Smart metering and consumption analytics enable reliable monitoring to support appropriate water pricing, while preserving privacy via data minimisation and aggregation rules.
1.4.5 Security and secure collaboration
RETRO-RO applies a resilience-by-design approach combining OT cybersecurity baseline controls (segmentation, role-based access, authenticated telemetry, integrity monitoring, secure logging) with continuity procedures and exercises. A secure INSPIRE-aligned data-sharing framework supports collaboration among stakeholders while explicitly mapping interdependencies with other critical infrastructures (energy and ICT).
1.5 Progress beyond the state of the art
RETRO-RO advances beyond mainstream 'smart water' approaches by combining: (i) bounded adaptation (explicit operating envelopes) rather than unconstrained optimisation; (ii) auditable, supervised decision support suitable for critical infrastructure procurement; (iii) effect-based safety validation (Biological Quality Gate) that complements chemical analytics under mixture uncertainty; (iv) coupling of asset-level retrofit with system-level digital twin, NbS buffering and secure data collaboration; and (v) explicit cyber-physical disturbance drills and fallback validation. Where appropriate, distributed ledger techniques may be assessed to strengthen audit-log integrity and enable verifiable evidence packages without exposing sensitive operational details.
2. Impact
RETRO-RO enables European utilities to maintain safe, affordable and low-carbon water services under climate extremes, emerging pollution and malicious threats. By validating a modular retrofit toolkit (digital twin + AI, hybrid grey-NbS buffers and cyber-physical security), the project delivers evidence-based replication assets-procurement specs, acceptance tests and good-practice guidelines-to accelerate uptake across utilities and regions.
2.1 Pathways to impact
RETRO-RO is impact-driven through deployable retrofit packages. It reduces climate and pollution vulnerability by enabling infrastructures to maintain safe service under extremes, reduces GHG emissions by stabilising and optimising energy and chemical use, and increases security and trust through auditable evidence and cyber-physical protection.
2.2 Key results and KPIs (baseline-comparative, falsifiable)
Service continuity under stress: quantified ability to maintain compliance and service during flood/drought-triggered variability and contamination scenarios (time-to-recover, service downtime).
Energy and GHG: reduction in specific energy consumption (kWh/m3) and variance; reduction in chemicals and associated CO2e; improved energy recovery performance where applicable.
Water safety: early-warning lead time and reduced false alarm rate for biological/contamination risk; effect-based toxicology, microbiological confirmation (including bacteria and viruses where relevant) and antimicrobial resistance screening within safety gates.
Upstream risk intelligence: coverage and performance of dynamic catchment-to-intake mapping for both groundwater-fed sources (springs and wellfields) and surface-water intakes (rivers, lakes and reservoirs), including alert hit rate and lead time (precision/recall), uncertainty reporting, time-to-sampling escalation, and reduction of intake contamination excursions and unplanned treatment disruptions attributable to early warning and operational playbooks.
Network resilience: leakage and infiltration/inflow detection/forecast performance; reduction of peak flows to plants through NbS buffering; reduction in overflow-related stress proxies.
Security posture: time-to-detect and time-to-recover in cyber-physical drills; number of implemented controls mapped to recognised OT security practices; integrity of audit logs.
Replication readiness: number of utilities/regions adopting the replication kit; procurement templates used; follower-site commitments.
2.3 Exploitation, replication and standardisation
The main exploitation route is infrastructure-oriented: utilities and integrators procure and deploy retrofit packages on existing assets. RETRO-RO delivers Key Exploitable Results (KERs) as procurement-ready artefacts:
Retrofit interface specifications (signals, sensors, actuation bounds, integration points).
Pre-registered stress-test scripts and acceptance criteria (commissioning-ready).
Auditable logging schema and data provenance model (tamper-evident evidence packages).
Digital twin components and AI models with deployment and maintenance guidance.
Catchment-to-intake risk mapping products: dynamic hotspot and pathway maps for springs, wellfields and surface-water intakes, plus interfaces that convert map outputs into operational triggers and sampling escalation protocols.
NbS integration patterns and monitoring protocols for utilities and municipalities.
Cybersecurity baseline package and incident-response playbooks for water OT contexts.
Good practice guidelines for resilience, redundancy, ecological, and social sustainability.
INSPIRE/FAIR-aligned data model and secure collaboration framework for stakeholders.
Standardisation and uptake will be supported through engagement with relevant European platforms and communities (e.g., Water4All, ICT4Water) and by mapping delivered interfaces and controls to widely used interoperability and security expectations for critical water infrastructures.
Complementarities will be actively managed by building on relevant results from previous Framework Programme projects, particularly those in the ICT4Water ecosystem, and by conducting periodic portfolio reviews to avoid overlaps with actions funded under the Mission "Restore our Ocean and Waters by 2030" and the European Union Mission on Adaptation to Climate Change. Where relevant, synergies with Horizon Europe Cluster 3 activities on critical infrastructure security will be sought.
2.4 Communication, dissemination and citizen/stakeholder engagement
RETRO-RO combines scientific dissemination with practitioner uptake: peer-reviewed publications; open repositories for non-sensitive assets; utility-focused webinars and training; and co-creation workshops with utilities, municipalities, regulators and citizen groups. Citizen-science activities are designed to be inclusive, privacy-preserving, and quality-assured. All non-sensitive in-situ data generated by the action will follow INSPIRE principles and will be made available through open access repositories where feasible (for example via Copernicus-compatible repositories), using a risk-based approach for security-sensitive operational data.
3. Implementation
3.1 Work plan overview
The work plan is organised into nine work packages (WPs). Each WP produces claim-ready outputs linked to acceptance tests, evidence packages, and measurable KPIs.
WP1 Coordination, quality management and ethics/security governance.
WP2 Retrofit module engineering (paired monitoring across source-water intake, product water and reject or concentrate streams; operating safety envelopes; operator-supervised recommendations; audit logging).
WP3 Nature-based Solutions as hydraulic and pollution buffers (design, permitting support, monitoring and integration patterns).
WP4 Digital twin and source-water risk intelligence (catchment-to-intake pollution mapping, EO assimilation, smart metering, predictive analytics).
WP5 Biological Quality Gate (effect-based and microbiological validation; threshold setting; safety confirmation under mixture uncertainty).
WP6 Secure collaboration and OT cybersecurity (baseline controls, threat modelling, incident response, disturbance drills).
WP7 Pilot implementation and stress-test ladder execution (ST1-ST4, matched baselines, evidence packages).
WP8 Replication, procurement packages and standardisation uptake (interfaces, acceptance tests, good practice guideline).
WP9 Communication, stakeholder engagement, citizen science and exploitation.
WP4 will deliver the integrated digital twin and the catchment-to-intake risk mapping as deployable assets, with explicit interfaces to retrofit control layers and utility supervision. Key tasks include:
Data harmonisation and metadata: INSPIRE-compliant geospatial layers, sensor data quality control, and provenance logging across sources (utilities, competent authorities, Earth observation).
Groundwater stream: recharge-zone delineation, vulnerability reasoning, and event-based mobilisation risk indicators for springs and wellfields.
Surface-water stream: flood and drought transport indicators for rivers, lakes and reservoirs, including pathway and hotspot ranking under extremes.
Operational integration: organization of maps into triggers and playbooks (intake switching/blending, sampling escalation, bounded treatment setpoint recommendations) and validation of lead time and hit rate against observed events.
3.2 Pilot strategy and validation logic
RETRO-RO will implement pilots that collectively cover both drinking water and wastewater contexts and that stress-test different failure modes (flood-driven inflow peaks, drought-driven concentration effects, emerging contaminants, and cyber-physical disturbances). Pilot details are provided as fill-in templates to be replaced with confirmed sites:
Pilot 1 (Drinking water, RO-based): [Utility/site, Country]. Stressors: drought conditions, salinity and temperature variability, and emerging contaminant pulses. Focus: energy stability, water-safety confirmation and auditable bounded actions. Catchment-to-intake risk mapping is operationally integrated: dynamic hotspot and pathway maps for springs, wellfields and nearby surface-water abstraction points provide early-warning lead time and trigger defined actions (intake switching or blending where applicable, targeted sampling escalation, and bounded adjustments to pre-treatment and reverse osmosis operation within predefined safety envelopes).
Pilot 2 (Wastewater / water reuse with advanced treatment): [Utility/site, Country]. Stressors: wet-weather inflow peaks, pollutant pulses, biological variability; focus on resilience and resource recovery readiness.
Pilot 3 (Urban network + NbS buffering): [Municipality/utility, Country]. Stressors: flood events and combined sewer impacts; focus on NbS buffering, infiltration/inflow analytics and community engagement.
Follower sites: [2-4 follower utilities/regions] adopting the replication kit without full pilot installation.
Validation follows a pre-registered stress-test ladder: TRL2-3 (indicator observability and calibration), TRL4 (integrated closed-loop tests with bounded actions and security controls), TRL5 (pilot operation with matched baseline comparison). All protocol deviations, missing data and operator overrides are recorded in the audit log and reported transparently.
Stress-test ladder (ST1-ST4) and acceptance evidence
ST1 (TRL2-3, lab): Controlled tests on retrofit components and observability indicators (RO performance, quality sentinels, control stability) under predefined drought-like and contamination-load variations. Acceptance: calibration/observability report; pass/fail against predefined KPI thresholds; complete traceability in the audit log.
ST2 (TRL3-4, pilot rig/skid): Integrated, closed-loop validation of the toolkit (sensing + decision logic + bounded actuation) under combined hydraulic shocks, variable salinity/temperature, and emerging-contaminant challenge mixtures. Acceptance: commissioning/acceptance-test scripts executed; safe operation within the defined operating envelope; reproducible evidence package.
ST3 (TRL4-5, in-situ pilots): Deployment in Pilot 1-3 for [X] months, capturing real climate-driven variability (drought/flood episodes) and operational disturbances. Acceptance: matched-baseline comparison demonstrating measurable improvements on energy/GHG, compliance/quality, resilience/continuity, and resource recovery readiness, without compromising safety.
3.3 Milestones and deliverables (high level)
3.4 Risk management
Key risks are managed through engineering controls, pre-registration of tests, and redundancy of evidence sources (sensors + lab anchoring + audit logs).
Upstream mapping uncertainty and data gaps: mitigate via multi-source fusion (EO + in-situ + historical records), explicit uncertainty bounds, conservative alert thresholds, and periodic recalibration against laboratory-confirmed samples.
Pilot access and operational constraints: mitigate via early site agreements, staged integration, and fallback test benches.
Sensor drift and false alarms: mitigate via calibration cycles, dual-sentinel confirmation logic, and confidence scoring.
Over-control or unsafe actions: mitigate via actions constrained within predefined operating envelopes, operator supervision rules and safe fallback states.
Data security and confidentiality: mitigate via OT/IT segmentation, access control, encryption, integrity checks and role-based sharing.
Citizen science data quality and privacy: mitigate via guided protocols, validation checks, data minimisation and aggregation.
NbS permitting and performance uncertainty: mitigate via early co-design with municipalities, conservative designs, and robust monitoring.
3.5 Ethics, data management and security
RETRO-RO will deliver a Data Management Plan (FAIR) with risk-based openness for security-sensitive operational data. Citizen participation will follow informed-consent and GDPR-compliant procedures. Security management includes threat modelling, incident response, and validation of safe fallback behaviour in disturbance drills.
References (indicative)
[To be completed with formatted references supporting: RO operating envelopes and critical flux, biofouling and effect-based monitoring, digital twins for water systems, flood/drought risk indicators, NbS performance evidence, and OT cybersecurity best practices.]
Environmental Institute - role
Project results are expected to contribute to all of the following expected outcomes:
water infrastructures are flexible enough to face changes in hydraulic flow and pollution load from emerging or yet unknown contaminants to ensure that access to water and sanitation is protected on a long-term, recovery of safe secondary resources is secured and greenhouse gas emissions are reduced and ecosystems are protected;
The proposed RO-based toolkit will be developed for and tested in three pilot case installations:
Production of drinking water via desalination of seawater;
Production of drinking water from surface water (direct intake from river/reservoir) in river basin impacted by frequent floods/droughts;
Reuse of wastewater for irrigation of agricultural crops (optional: recharge of aquifer used for production of drinking water).
The source water in the first two pilot cases and wastewater influent/effluent will be thoroughly characterized for presence of emerging contaminants using novel state-of-the-art wide-scope target screening (>2,800 substances) and suspect screening (>95,000 substances) analytical techniques [REF Alygizakis et al.; link to EMPODAT and DSFP]. (Eco)toxicity threshold values for of all suspect chemicals are available in the NORMAN Ecotoxicology Database [link]; human toxicity threshold values will be taken from the EU Drinking Water Directive (DWD; REF) and/or derived using the novel ToxAI tool [REF]. Detected contaminants will undergo risk assessment using NORMAN Prioritisation Framework [Dulio et al, 2024] and a list of typical priority contaminants will be developed for each pilot installation. Non-target screening (NTS) by liquid (LC) and gas chromatography (GC) high-resolution mass spectrometry (HRMS) techniques allows for detection of yet unknown contaminants or their retrospective screening. In case of their permanent presence or increase of their signals an effort will be made to identify them and include in the list of target compounds. Exceedance of toxicity threshold value of any of the above substances will be automatically reported to the central AI-based operational system with clearly defined SOPs for follow up actions.
A parallel 'safety net' analyses of source water and wastewater effluents by a battery of well-established and robust in vitro bioassays addressing a wide-range of specific modes of action (ERα, GR, Anti-AR, PPARγ, PAH, PXR, Nrf2) will ensure that toxic effects of emerging contaminants not detected above are captured. The effect-based data will be linked with chemical occurrence data. Several indicator values (e.g. toxic units, effect concentrations), considering concentrations of identified/determined substances, their mode of action and normalized toxic properties, will be used to explain the toxicity signal of each of the bioassays and identify chemicals responsible for a major part of the mixture toxicity (toxicity drivers). A new unique BioActivity Database module of the NDS [link] will be used for this purpose. An automated response plan will be for operators and authorities will be initiated (by the central AI-powered control system) in case of exceeding the Effect-based Trigger (EBT) values by individual bioassays [Alygizakis et al.]. All data will be available in harmonised formats in the open-access NORMAN Bioassays Monitoring Database [link].
A modular installation of continuous in vivo bioassays (zebra fish and molluscs) will be tested in pilot case 2 [REF RIWA Meuse river, NL]. Reduced agility of fish or lesser frequency of opening of mollusc shells are automatically detected and trigger an alarm. An automated storage of 50 l source water is then initiated for follow up analyses to identify the cause. The intake of source water will be diverted until the alert is explained/source water quality is back to normal.
The EU Water Reuse Directive [REF] establishes harmonised minimum requirements for the safe reuse of treated urban wastewater, primarily for agricultural irrigation, to address water scarcity and promote circular economy practices. All parameters as listed in the directive (E. coli, BOD5, TSS, turbidity, Legionella spp, Intestinal nematodes, Total coliphage, Clostridium perfringens spore etc.) with appropriate frequencies will be included in the pilot case 3. As recommended by the WRD (and in line with the latest update of the UWWTD), the reused wastewater will also be tested for heavy metals, pesticides, disinfection by-products, pharmaceuticals, other substances of emerging concern, including micropollutants and microplastics and anti-microbial resistance. The well-characterised wastewater effluent ensures that application of reused wastewater will not cause additional harm to aquatic ecosystem in cases of run off to rivers. At present, there are no threshold values for chemicals indicating their adverse effects to soil and damage to terrestrial ecosystems. Here, Environmental Institute will bring in the latest know how from the on-going Horizon TerraChem project [REF].
The pattern of chemical pollution in source water is usually stable in long-term [REF Rhine/Meuse DW installations] and low frequency of measurements (weekly/monthly) is sufficient. An AI-based system for statistical processing of the concentration data and their trends will be installed, and any deviation will be automatically reported to consider technology-based actions (diversion of the flow, change of treatment parameters etc.). An event-based (change in hydraulic flow/pollution load - floods/droughts; accidental pollution warning) triggers will be defined to increase the frequency of measurements and initiate follow up actions by operators/authorities.
water infrastructures have integrated digital solutions (e.g., smart sensors, IoT, digital twins and artificial intelligence) as well as citizen science to optimally operate in changing conditions from climate or pollution pressures, facilitate appropriate water pricing based on reliable monitoring of water consumption, favour recovery of material and limit greenhouse gas emissions;
An Early-Warning System (EWS) reflecting pollution pressures will be incorporated into the central AI-based control system. The prototype co-developed by Environmental Institute within the NORMAN network [REF], TerraChem [REF] and PARC [Partnership for the Assessment of Risks from Chemicals; all EU MS, REF] will be adjusted to the conditions of pilot case studies. The EWS provides automatically signals when concentrations of emerging contaminants, unknown contaminants, exceedances of EBTs, excessive presence of AMR genes, microplastics etc. are of concern and triggers follow up technology adjustments or actions by operators.
To tackle the drinking water pollution, EU DWD includes the risk-based 'source to tap' approach. Regarding surface water, EU Water Framework Directive (WFD; REF) and updated UWWTD promote the 'polluter pays' principle. However, there is never sufficient time, funds and programmes to investigate pollution in large geographical areas (river basins, coastal zones). The project will therefore test citizen science approach - when the central AI-based control system detects any anomaly in source water quality a team of volunteers (e.g. from secondary schools) will immediately sample surface water with simple sampling device (stir-bar; c. 2 cm glass rod put in c. 500 ml water on magnetic stirrer for 15-20 min.). After extraction, the stir-bar is put into envelope and sent in a colling box to a laboratory for follow up analysis of wide range chemical contaminants by GC-HRMS. The whole process is quality controlled by addition of isotopically labelled internal standards and the data can be used to track the source of pollution [REF JDS5 EC JRC].
water infrastructures have incorporated the necessary tools and protection to avoid cyber and/or terrorist attacks to ensure their resilience against malicious behaviour;
Drinking water infrastructures are vulnerable to terrorist attacks [REFs]. In this project we shall focus on preventing (i) cyber-attacks, where remote computers attack valves, pumps and chemical processing equipment though computer-based controls, and (ii) chemical and biological attacks, in which terrorists introduce water-soluble biological or chemical contaminants into a publicly accessible city water supply. The comprehensive suspect screening of >95,000 substances (cf. above) already include numerous major organic chemicals (e.g. poisons, toxic pesticides, pharmaceuticals, chemical warfare agents, biotoxins). The list will be extended at the beginning of the project for the latest known lists of substances suitable for such purposes by extensive AI-powered literature search. The data will be stored in the NORMAN Substance Database [link].
The attack is usually aimed at the most vulnerable part of the water supply system - pipeline from the source water. The pollution pattern (by thousands of chemicals) will be reduced to a list of potential threat chemicals, which will be measured in source water and at the point of first barrier in the treatment technology. The minimum required frequencies will be provided in the SOP. Any difference between the two signals will trigger immediate alarm and diverting flow of source water using the central AI control system.
Similarly, we shall establish a list of known microbiological parameters, including bacteria, viruses, parasites, which will be included in regular monitoring schemes at two points - source water and first barrier in the treatment technology.
new or renewed water infrastructures are designed following 'good practices' to maximise system resilience, build redundancy, and ensure ecological and social sustainability;
The new or renewed drinking water and wastewater treatment infrastructures must comply with the requirements of EU environmental policy, considering not only the DWD, WFD, WRD, UWWTD but also GWD, IED, EU Chemicals Strategy, Soil Monitoring Law, One Health approach and EU Biodiversity Strategy. In the proposed project we aim at the 'best practices' to maximise system resilience in terms of tackling the reduction of 'risk of universe of chemicals', AMR, microplastics and microbiology threats in one AI-controlled system using modular monitoring and treatment technology components. Reduction or elimination of these threats will ensure ecological (pollution of source waters) and social (acceptance of wholesomeness of drinking water and using reused wastewater in circular economy) sustainability.
water infrastructures are better designed thanks to improved predictions, robust assessment of impacts and implementation of appropriate mitigation measures due to advances on the understanding of how, when and where floods and droughts occur.
The WFD provides a suitable framework to address water scarcity and drought. Water quantity management is also addressed through other EU regulations: (i) the Regulation on minimum requirements for water reuse for agricultural irrigation, that establishes new rules to stimulate and facilitate water reuse in the EU; (ii) the Recast of the EU Drinking Water Directive, that addresses leakage in the water supply networks. Further supporting water quantity management are the Commission proposals to revise the Urban wastewater treatment directive (UWWTD) and the Industrial Emissions Directive. Water scarcity and droughts are recognised as a priority in the European Green Deal and are reflected as such in several major European strategies: (i) the 2021 EU Strategy on Adaptation to Climate Change and (ii) the 2020 Circular Economy Action Plan and the Biodiversity Strategy for 2030.
The project will systematically screen the Flood Risk Management Plans and Drought Management Plans already developed in the pilot case basins with focus on the assessment of impacts and mitigation measures already accepted at the national/basin scale.
To achieve the expected outcomes, proposals should address some or all the elements below:
develop and integrate modular processes and tools to improve the adaptability of drinking water and wastewater infrastructures to emerging pollutions and effects of climate change;
The project will develop and integrate modular processes and tools to improve infrastructures to emerging pollution while reflecting floods and droughts effects.
integrate Nature-based Solutions infrastructures to reduce the carbon footprint of water infrastructure, better manage water flows and pollutants entering sewers, and support biodiversity;
enhance the use of digital solutions, new monitoring techniques, Earth observation tools, digital twin technology and artificial intelligence, including for predictive analytics, within drinking water and wastewater infrastructures to optimise their operation, anticipate infrastructure challenges and pollution and improve their efficiency and resilience, addressing leakages, infiltration, energy consumption, recovery of materials and carbon footprint;
The project will introduce state-of-the-art novel monitoring techniques for thousands of emerging contaminants (GC- and LC-HRMS), microplastics (by both Py-GC-MS and FTIR), EBM and AMR (cf. above). Earth observation tools will be used to feed the flood and draught information in the pilot basins. The AI-powered digital twin technology will simulate the performance of pilot technologies and report on any deviation with proposal of immediate corrections. A critical mass of data on chemical and microbiological pollutants will be collected in the first part of the project, which will allow for predictive analytics by analysis of patterns of pollution instead of individual substances.
develop robust data sharing framework to promote secure collaboration among stakeholders and identify interdependencies with other critical infrastructures in a resilience-based approach;
Environmental Institute has extensive experience in development and maintenance of data sharing framework at the EU scale [REF NORMAN] supplying stakeholders with various data interdependencies required for implementation of EU policies and regulations in the environmental sector. This holds specifically for resilience-based approach, where different data in differing formats addressing various directives must be brought together (emerging substances, NTS, microbiology, AMR, microplastics, ecotoxicology, human toxicology, flood risk and drought maps, INSPIRE, big data from critical infrastructures via on-line measurements, etc.).
develop tools, approaches and procedures to protect both the physical and digital water infrastructures against malicious attacks.
Addressed above.
Proposals should also seek to contribute to the further development of existing observing platforms and initiatives, including to the evolution of Copernicus services and the future EU Digital Twin on freshwater and Destination Earth. It should also contribute to define the bases for a FAIR (Findable, Accessible, Interoperable and Re-usable) sharing of data in the water sector in collaboration with the initiative conducted by the co-funded partnership Water Security for the Planet (Water4All).
All in-situ data collected through actions funded from this call should follow INSPIRE principles and be available through open access repositories (e.g. Copernicus). supported by the European Commission.
All data obtained in the project will be openly accessible in the Name of the Project Data Management System using FAIR principles. The data will be shared in the water sector using NORMAN Database System (NDS; link) developed by the NORMAN network of more than 100 organisations in Europe and beyond dealing with all aspects of emerging contaminants in environment since 2005. Harmonised data formats allowing for interoperability and reusability of findable and accessible open access data are well in place. The NDS is being developed and maintained by the project partner Environmental Institute and it is well recognized by the European Commission and its agencies as a part of the EU Common Data Platform for Chemicals [REF]. The same approach is used by several Water4All projects (to be approved in the second round).
Project overview (elevator pitch) / RETRO-RO turns today's drinking water and wastewater infrastructures into the water systems of the future-without rebuilding them. We deliver a modular, plug-and-play retrofit toolkit that keeps services safe and compliant under floods, droughts and rapidly changing hydraulic and pollution loads, including emerging and unknown contaminants. A secure digital twin + AI layer fuses IoT/SCADA, lab data, smart metering and Earth-observation/climate indicators to move from reactive to predictive operations-cutting energy use and GHG emissions while protecting service continuity. We combine "grey" upgrades with Nature-based Solutions that attenuate peak inflows and pollutant pulses and support biodiversity. Cyber-physical security-by-design hardens critical assets and accelerates detection and recovery from malicious attacks. Validated in real pilots through a staged, pre-registered test programme and matched baselines, RETRO-RO delivers replication packages (procurement specs, acceptance tests, SOPs, cybersecurity baseline and FAIR/INSPIRE-aligned data models) to scale uptake across Europe. / Headline impact (validated in pilots): / Cut energy use and GHG emissions per m³ while maintaining compliance. / Improve resilience and service continuity under flood/drought and load shocks. / Detect and respond faster to contamination and cyber/physical incidents.
Topic requirement | RETRO-RO contribution | Main WPs / outputs
Outcome 1 - Flexibility under flow and pollution variability (incl. emerging/unknown contaminants); safe resource recovery; GHG reduction; ecosystem protection | Modular retrofit processes (RO-centred demonstrator) plus NbS buffering; stress-tested operating envelopes; effect-based Biological Quality Gate; energy/chemical reduction and compliance-first routing. | WP3, WP4, WP8; Deliverables: retrofit specs, envelope protocol, safety validation report, KPI dataset
Outcome 2 - Integrated digital solutions (sensors/IoT, digital twins, AI) + citizen science; consumption monitoring for pricing; materials recovery; GHG reduction | Digital twin + AI predictive analytics integrating sensors, smart meters, citizen science signals and Copernicus indicators; decision-support for pricing-relevant consumption monitoring and recovery readiness. | WP2, WP5, WP8, WP9; Deliverables: twin platform, AI models, citizen science protocol, metering analytics guide
Outcome 3 - Tools and protection against cyber and/or terrorist attacks | Resilience-by-design security baseline for OT and data exchange; disturbance drills and safe fallback; integrity and logging; incident-response playbooks. | WP7, WP6, WP8; Deliverables: security architecture, test reports, incident response SOP
Outcome 4 - Good practices for new/renewed infrastructures: resilience, redundancy, ecological and social sustainability | Good-practice guideline and procurement-ready replication packages; redundancy patterns; ecological/social sustainability criteria; commissioning/acceptance procedures. | WP9; Deliverables: guideline, procurement templates, replication kit
Outcome 5 - Better design through flood/drought prediction, robust impact assessment and mitigation measures | Flood/drought indicators and scenarios integrated in risk assessment and stress-test design; EO-based triggers; mitigation options for asset and network operations. | WP5, WP2, WP8; Deliverables: climate stress scenario library, forecast/trigger evaluation, mitigation handbook
WP | Title | Main outputs
WP1 | Project management, quality assurance and ethics/security compliance | Governance, risk and quality plan; deliverable acceptance logic; ethics and data protection oversight; security management.
WP2 | Baselines, requirements and co-creation (utilities, authorities, citizens) | Baseline assessment; user requirements; citizen science and pricing requirements; pilot protocols; KPI baselines.
WP3 | Modular retrofit processes for treatment assets (RO-centered demonstrator) | Guard/Mirror sensing; envelope methodology; bounded actuation; concentrate routing; Biological Quality Gate integration.
WP4 | Nature-based Solutions integration and monitoring | NbS design and deployment patterns; hydraulic/pollution buffering; biodiversity and carbon footprint assessment.
WP5 | Digital twin, predictive analytics, Earth observation and source-water risk mapping integration | Twin architecture; EO indicators; flood/drought triggers; predictive analytics for leakage/infiltration/inflow and energy/GHG; decision-support outputs.; source-water risk maps and trigger library; microbial hazard indices
WP6 | Secure data-sharing framework and interdependency mapping | INSPIRE/FAIR-aligned data model; secure collaboration; interdependencies with energy/ICT; governance and access rules.
WP7 | Cyber-physical security and continuity procedures | Threat modelling; OT security controls; secure logging; incident response; disturbance drills and safe fallback validation.
WP8 | Pilots, stress-test ladder and validation | TRL2-3 calibration; TRL4 integrated tests; TRL5 pilot operation; matched baseline comparisons; evidence packages and KPI datasets.
WP9 | Replication, good practices, exploitation and communication | Replication kit; procurement templates; good-practice guideline; training and dissemination; standardisation liaison.
Milestone | Description
MS1 | Consortium and pilot protocols frozen; ethics/security plan approved (M3)
MS2 | Baseline datasets and requirements validated; citizen science protocol ready (M6)
MS3 | Retrofit modules integrated; envelope and safety gates calibrated (M12)
MS4 | Digital twin operational with EO triggers; secure data-sharing framework deployed (M18)
MS5 | Pilots completed; KPI evidence packages validated; replication kit and good-practice guideline released (M36)