The HARMLESS partners developed a modular, stepwise approach to safely and sustainably innovating AdMa that aligns with the industry's multi-phase product development process. The early stages provide general safety guidance based on the available information about AdMa and its intended use. Later stages identify critical knowledge gaps and recommend new approach methodologies (NAMs) that are cost-efficient and can acquire a harmonized dataset for environmental and health risk assessments. These assessments are conducted in the context of known risks from reference materials, using advanced computational modeling. In the latest stages, high-throughput toxicity screening and multi-omics analysis provide the most comprehensive risk assessment.
Bioinformatics and multiscale computational models, including quantitative structure–activity relationship (QSAR) models, were used to predict in vivo hazard potential for over 70 reference and case study materials from NAM data and to guide the design of safer materials. To cluster these materials in a more bio-mechanism-aware manner, HARMESS researchers mapped the community-accepted network of adverse outcome pathways (AOPs) (AOPwiki.org).
HARMLESS created advanced tools to manage and analyze big data sets of millions of data points for the over 70 materials in the eNanoMapper database, ensuring data are FAIR (Findable, Accessible, Interoperable, and Reusable). Automated data integration and visualization within eNanoMapper improved transparency and accessibility for regulatory and industrial stakeholders.
The tools were integrated into a user-friendly Decision Support System (DSS) to help companies apply Safe-and-Sustainable-by-Design (SSbD) principles during product innovation. Multiple industrial case studies demonstrated real-world applicability of the DSS with refined tools and workflows. An Early Warning System (EWS) was also developed to identify potential risks early in material innovation, providing timely input for regulators.
Main results:
• Mode-of-Action and AOP-based Integrated Approaches to Testing and Assessment (IATAs) were successfully established for a broad range of advanced and conventional materials.
• More than 40 NAMs providing data for ca. 200 parameters, including quantum-mechanical descriptors, were refined within this project.
• Large, multi-level datasets for health and environmental risk assessment were generated and made FAIR by uploading them to the eNanoMapper platform (enanomapper.adma.ai/projects/harmless; open access from November 2026).
• High-throughput in vitro toxicity scoring and transcriptomics analyses provided new mechanistic insights and enabled grouping and safety assessment of materials.
• The developed over 100 predictive models (e.g. QSAR) linking NAMs to the toxicity response of animals are essential for leveraging NAMs in health-related risk assessment.
• The user-friendly DSS offers a practical, stepwise workflow aligned with the SSbD strategy of the European Commission (DOI: 10.2760/879069) to support safer material design decisions from early- to late-stage product development.
• The HARMLESS Early Warning System was shown to align with international frameworks (e.g. OECD Early4AdMa) and demonstrated its value for improved regulatory preparedness through collaborative workshops with OECD on two HARMLESS case study materials.
• Extensive stakeholder engagement ensured that tools and methods are compatible with real-world industry needs and regulatory requirements.
• The DSS was effective in all four industry case studies, from ideation to laboratory phases leading to varied decisions: Stopping the project before scale-up (automotive catalysts by perovskites), renewed ideation (plant protection by imogolites), and continued marketing with proven safety and sustainability (paints with colloidal silica and insulation by aerogel-mats).
