Infra-red spectroscopy and AI help accelerate binder performance predictions – 12th Eurobitume-Academia webinar
On 16th April 2025, more than 30 people attended Eurobitume’s 12th online academia webinar, which explored how infra-red spectroscopy, AI (artificial intelligence), and machine learning can speed up the prediction of binder performance.
Presenting work carried out at Aalto University in Finland, Professor Augusto Cannone Falchetto described the evolution of new characterisation methods – infra-red spectroscopy and AI / machine learning – that can be used to predict binder performance faster than traditional DSR / rheology plus ageing tests.
Professor Cannone Falchetto stated that the ultimate outcome of this ground-breaking work is to be able to use spectroscopy to predict the rheological properties of binders much more rapidly than carrying out the test.
He explained the development of “principal component analysis”, a data reduction / statistical method designed to simplify multiple data sets, as a predictive tool and said they used different machine learning models to predict rheological properties.
Augusto concluded by stating that the final phase of their exhaustive research involved the analysis of almost 26,000 data points.
After the presentation, there was a lively discussion and Q&A session.
Eurobitume members will soon be able to access a recording of this academia webinar on the Training / Academia Webinars page of the Eurobitume Intranet site, where recordings of previous webinars are accessible.
The next Eurobitume-Academia webinar will be delivered by Manolis Romanias, Assistant Professor at the University of Lille IMT Nord Europe, on September 4th, 2025, from 10:00 to 11:30 CEST.
