Post by Quandela

Four partners, one industrial materials challenge 🧪   TotalEnergies, MBDA, Alysophil and Quandela partnered through the PolyT project to explore whether quantum machine learning could support the assessment of polymer thermal stability.   Predicting how polymers respond to heat is essential for developing materials that meet specific performance and safety requirements. Yet polymer structures are complex, experimental datasets are often limited, and evaluating large numbers of formulations can require significant time and resources.   During Phase 1b, the team developed a hybrid workflow in which classical data processing converted polymer information into a compact representation, which was then analysed by a quantum classifier running on Quandela’s Ascella photonic quantum processor.   ⚙️ Across five hardware runs, the model achieved around 84–85% classification accuracy, with consistent results, performance close to noise-free simulations and outcomes competitive with the classical benchmarks evaluated.   The study shows that an industrially relevant materials-classification problem can be executed consistently on real photonic hardware and integrated into a broader machine-learning workflow.   For industry, this provides a practical basis for exploring how quantum computing could support materials R&D, particularly when data is scarce and structures are complex. While still an early step, the project establishes a technical foundation for further research as hardware, algorithms and datasets improve.   ✍️ Read the blog here: https://lnkd.in/e5s8uacm Arno Ricou Brian Ventura Aleksandrina S. Vassilis Apostolou Philippe Robin Bogdan Penkovsky Jérémie Messud Jean-Patrick MASCOMERE Léo Dumas Vincent GARDON Denis Gardin Marc Bouchez Pascale Senellart Jean Senellart Valerian Giesz Niccolo Somaschi #QuantumComputing #QuantumMachineLearning #Photonics #MaterialsScience