Tassos Papadiamantis is a senior scientist at the Entelos Institute (Cyprus). Within SSbD4CheM, he is leading the work package dealing with the computer aided (re)design approach.
Tell us a bit about yourself. What is your area of expertise?
Tassos Papadiamantis: My expertise is at the interface of nanomaterials safety, data and AI governance and management, and responsible predictive modelling. On the data side, I work on making nanosafety, materials, and chemicals data FAIR. On the modelling side, I develop responsible AI and machine-learning approaches to predict how materials behave, e.g., read-across models for ζ-potential and stability. I also work on developing ethical AI assessment frameworks applicable in materials and chemicals research. I bring this work into the regulatory space, mainly through projects with the European Union Observatory for Nanomaterials and EFSA, and publicly and commercially funded projects, supporting REACH and CLP compliance. This is also what I bring into SSbD4CheM and WP2, data-driven modelling, responsible AI, and FAIR data practices feeding directly into safe and sustainable design.
How does your specific work package “Computer aided (re)design approach” contribute to the project?
TP: WP2 provides the computer-aided (re)design layer of SSbD4CheM. Our role is to bring computational tools into the SSbD framework so that material selection and re-design decisions can be made early, before going to the lab. We combine physics-based methods and data-driven workflows that predict properties, behaviour, fate and transport of candidate materials. From these results we also identify the data gaps that need experimental work, which then guides what is actually measured in the project. On top of that, we develop a data-driven LCA estimation framework, so even for materials that do not yet have full inventory data we can give an estimation on their environmental impact. In practice, WP2 is the layer that turns SSbD from a principle into something computable and iterative.
What is the most exciting thing about the activities in your work package?
TP: For me the most interesting part is the loop we are building between models and experiments. The data-driven workflows we set up are not delivered once and frozen. They can be refined with new experimental data and the refined models can then feed the next design cycle. Combined with the LCA estimation step, this lets us look at safety, performance and lifecycle impact of a material that may not exist yet in the lab, which is something quite hard to do today. And we get to test this on three very different case studies, i.e., renewable composites for automotive, PFAS-free coatings for textiles, and cellulose nanofibers replacing plastic microbeads in cosmetics.
- Tassos Papadiamantis
Senior Scientist at Entelos Institute“In SSbD4CheM, we bring computational tools into safe and sustainable design, enabling early, data-driven decisions before materials reach the lab. By combining AI, predictive modelling and lifecycle estimation, we create an iterative loop between models and experiments, helping industry and regulators assess safety, performance and environmental impact even for materials that do not yet exist.”
From your point of view, who can benefit the most from the project?
TP: The most direct beneficiaries are the chemical and material producers and the downstream industries in automotive, textile and cosmetics. These are the stakeholders that carry the cost and the risk when a material turns out, later in development, to be unsafe, non-compliant, or with a poor environmental profile. With the WP2 tools they can take SSbD decisions earlier, with fewer experiments and with a clearer picture of the full lifecycle. Beyond industry, regulators and standardisation bodies also benefit, because the methods are validated following OECD principles and reported in formats they can directly use (for example QMRF for the predictive models). And in the end consumers benefit too, since products reaching the market will have been screened for safety and sustainability before scale-up, not after.
