April was a dynamic month for the PINK project, marked by a strong presence at the European Materials Modelling Council (EMMC) 2025 conference. As part of the project’s ongoing commitment to collaboration and knowledge exchange within the materials modelling community, representatives from the PINK consortium engaged in the conference through a diverse range of contributions.
The PINK team was actively involved in multiple formats: coordinating and moderating the session on sustainability with a dedicated panel discussion, delivering oral presentations on key project insights, and showcasing research through scientific posters. These activities underscored the consortium’s multifaceted expertise and its role in shaping current discussions in the field of materials modelling and digitalisation.
The following chapters provide a closer look at each of these contributions, highlighting the topics covered, the experts involved, and the impact of their engagement within the broader EMMC 2025 framework. One of the highlights was the PINK conference session “Sustainability as innovation driver for a zero-pollution and climate-neutral future” which was organised by Steffi Friedrichs (AcumenIST) and Thomas Exner (Seven Past Nine) on the 10th April.
During the session Martin Himly (from consortium partner PLUS) gave an invited talk with the title “Computational model integration across SSbD dimensions on advanced materials and chemicals: development driven by industry and innovators needs” he presented PINK and its cutting-edge initiative developing integrated computational tools to support Safe-and-Sustainable-by-Design (SSbD) chemicals and advanced materials. Tailored to industry needs, PINK provides an open innovation platform (PINKISH) specifically addressing the complex challenge of balancing safety, sustainability, functionality, and cost-efficiency in the development of advanced materials and chemicals and, thus, directly supports Europe’s green and digital transition goals.
Picture: Copyright EMMC 2025
Philip Doganis (from NTUA) presented JAQPOT in his talk, a computational cloud platform developed to support the design of chemicals and materials within a SSbD framework. Jaqpot enables the deployment and management of machine learning (ML) and other computational models such as Physiologically Based Kinetic (PBK) models without the need for local computational infrastructure. It offers a user-friendly interface and robust application programming interfaces (APIs) for handling custom and pre-existing models, enhanced by advanced pre-processing and featurisation methods that improve model accuracy and interpretability. The platform supports popular ML frameworks including Torch and Scikit-Learn and is compatible with ONNX runtime, while a dedicated Python client (Jaqpotpy) facilitates seamless workflow automation. Additionally, ongoing expansions include integration with generative learning techniques and large language models to further promote SSbD principles. The platform thus streamlines key aspects of the ML lifecycle such as collaboration, data security, and model performance, allowing materials scientists to concentrate on innovation.
These two presentations from PINK consortial members with other oral contributions laid ground for the following panel discussion “Sustainability through the Materials Modelling and Data Ecosystem of the Future”: Moderated by Steffi Friedrichs, this engaging panel brought together experts from academia and industry to explore how sustainability can be achieved through collaborative innovation, data-sharing frameworks, and interoperable infrastructures in materials science.
Steffi Friedrichs opened with a powerful reminder: “Innovation needs to be reliable and resilient, to be dependable both by industry, governments and the public.” She called for a Research & Innovation ecosystem rooted in shared responsibility and trust, where collaboration produces results greater than the sum of its parts.
Eva-Kathrin Schillinger (IAM-I) highlighted the complexity of SSbD, emphasizing that sustainability must be assessed not just at the materials level, but also across components, systems, manufacturing, and processing stages: “We have to ultimately arrive at a (data-driven) approach that allows us to fail fast and fail cheap.”
Picture: Copyright EMMC 2025
David Rouquie (ECETOC and Bayer Crop Science) outlined the necessary building blocks for safe and sustainable access to data:
- Federated data systems
- Trusted data services
- Robust governance models
- Clear legal frameworks and strong incentives
Masahiko Demura (National Institute for Materials Science, Japan) advocated for openness and flexibility in data structures, promoting federated learning as a powerful tool to maintain sovereignty while enabling collaboration: “The use of federated data instils reliability; it is the quality of the data structure that decides if the data is correctly described to ultimately enable truly data-driven systems.”
Simon Stier (Fraunhofer ISC) underscored the need to embed sustainability early in the innovation pipeline using interoperable digital tools. “The main value of the task at hand lies in the model: this combines the product design with the process design.”
Key Discussion Points in the further discussion were:
- How can materials be designed for sustainability amid geopolitical change ?
- What infrastructure is needed to support federated, secure access to modelling and data sharing ?
- How can industry be incentivized to share data and collaborate more openly ?
This panel underscored a vital truth: sustainability in materials innovation depends not only on technical excellence but also on digital trust, regulatory clarity, and cross-sector collaboration.
The PINK appearance was rounded off with posters by Francesca Bleken (from SINTEF) and Francesco Mercuri (CNR-DAIMON). Francesca presented her and her coworkers’ results in form of a methodology for enabling modular and machine-interoperable workflows through the FAIR documentation of datasets, models, and related resources using semantic technologies. The approach is grounded in the European Materials Modelling Ontology (EMMO), ensuring that all documented entities correspond to real-world instances at the time of registration. Abstract resources such as models and datasets are represented as classes, while their executions and instantiations are captured as individuals. Interoperability is achieved by aligning input and output data models within the DLite framework, facilitating integration across domains and enabling automated workflow discovery through ontological reasoning and graph traversal. The methodology further incorporates the DCAT standard to enhance accessibility of dataset metadata, and is supported by tools such as Tripper and DLite, developed within several EU and Norwegian-funded research projects.
Francesco presented in his poster the DAIMON Team, based at CNR-ISMN in Bologna, as a multidisciplinary research group with expertise in digitalisation, artificial intelligence (AI), advanced simulations, and high-performance computing (HPC). They apply state-of-the-art computational infrastructures to address complex challenges across a broad range of applications and value chains, including advanced manufacturing, healthcare, energy, ICT, quantum technologies, agri-food, mobility, and more. This experience is provided to the PINK project in the form of automated workflow generation for execution of complex multiscale modelling tasks to predict SSbD indicators as well as performing such calculations for the PINK Developmental Case Studies predicting characteristics of advanced materials and their interactions with the environment.
