Early Warning System

The European Green Deal aims to protect human health and the environment while moving toward a toxic-free future. In support of this goal, PARC is developing an Early Warning System (EWS) to proactively identify emerging chemical risks and support timely regulatory interventions across the EU. 

The PARC EWS aims to establish a systematic, permanent, and coordinated framework for detecting and addressing chemical threats before they become widespread. It will directly contribute to the development of the European Commission’s Early Warning and Action System (EWAS) and embodies the vision of a systematic, permanent, and coordinated EU-wide framework for identifying and managing chemical risks before they escalate.

By leveraging big data, multi-source integration, and artificial intelligence (AI), the PARC EWS will go beyond expert-driven assessments, creating a dynamic and automated signal detection system that strengthens regulatory preparedness and environmental protection.

The PARC EWS will follow a structured methodology for identifying and validating early warning signals of chemical risks. This process will consist of three key stages:

  • Screening and filtering of signals (e.g. non-target screening (NTS), effect-based monitoring (EBM), effect directed analysis (EDA), literature, omics, patent databases).
  • Confirmatory check for legislation and measures.
  • Acquisition of information and signal amplification.

This systematic approach enables the EWS to track initial data signals, e.g., from mass spectrometry, omics, or patent analyses, through to validation, refinement, and ultimately supporting decision-making and policy responses. While feedback loops may not be explicitly detailed in all workflows, they are integral to the system’s continuous evolution.

Identification framework and prioritisation 

A core function of the EWS is the systematic identification and prioritisation of substances to ensure effective monitoring across environmental, human health, and product safety domains, while handling high-throughput data. To achieve this, the EWS system will:

  • mine international chemical databases (e.g., REACH, GHS, PubChem) to gather structural and usage information on substances,
  • apply AI and machine learning algorithms to predict toxicity directly from raw data, such as MS2 spectra,
  • integrate metadata, structural alerts, and exposure reconstructions to identify substances of emerging concern,
  • develop a scoring system using multicriteria decision analysis to support transparent and systematic prioritisation of substances. 

Computational platform development and integration

The PARC EWS will function within a sophisticated computational environment that:

  • enable real-time signal detection from big data sources across environmental and human health domains,
  • integrate advanced tools such as QSARs, read-across methods, bioinformatics models, exposure estimation tools, and Adverse Outcome Pathway (AOP) frameworks, to enhance risk assessment,
  • promotes FAIR data practices to ensure regulatory usability and interoperability,
  • offers a user-friendly interface, including a centralised dashboard and interactive “EWS wizard” to guide stakeholders through the system. 

Policy interface and regulatory alignment

The PARC EWS is being developed in close alignment with EU regulatory frameworks, and aims to incorporate improved chemical risk management for non-regulated compounds, including pharmaceuticals, PFAS, and polymers. Ongoing efforts focus on harmonising methods and guiding the use of innovative data types (e.g., NTS, EDA, EBM) in a regulatory context.

PARC EWS validation and case studies

To validate the practical use of the PARC EWS, real-world case studies are underway:

  • Suspended particulate matter (SPM) samples from German rivers are analysed using NTS, EBM and EDA to identify hazardous substances.
  • Fish samples from Greek river are screened for toxic chemicals related to food safety using NTS, EBM and risk assessment.

These case studies help refine EWS methodologies, particularly regarding data interpretation and tool integration across different exposure contexts.

Knowledge sharing and future developments

The first EWS prototype is currently under development, with a functional version expected by late 2025. The ultimate aim is to deliver a comprehensive, AI-enabled EWS that enables regulators, researchers, industry, and citizens to anticipate and mitigate emerging chemical risks, supporting a safer and more sustainable future for Europe.