10121913Self-learning AI Copilots to Enable Personalised Medicine ManufacturingActiveCollaborative R&DInnovate UKThis project seeks to develop a data-efficient vision-based AI process monitoring system to seamlessly assure additive manufacturing (AM) part quality in real-time, thereby accelerating AM adoption including in our case study of personalised pharmaceuticals. Our key innovation will address a fundamental challenge with AI systems: they typically need a lot of data, which can be difficult in manufacturing. As Matta moves into higher-value processes where data is more expensive to gather yet stakes are higher, we will need to train our systems with less data. This project will therefore pioneer new AI methods to reduce the need for costly experimental training data while enhancing the performance of its AI systems. Pharmaceuticals are an ideal case study because AM has demonstrated great potential to transform medicine by personalisation and rapid prototyping of dosage forms at the point of care. However, strict quality control and specific requirements for every produced batch are required, when releasing pharmaceutical products to patients. This currently prevents the wide adoption of 3D printed pharmaceuticals. Hence, an agile manufacturing technology that can be supervised by artificial intelligence is of critical value to enable applying AM in the pharmaceutical field. The data-efficient AI copilots developed in this project may thereby help vastly more patients access personalised 3D printed medicines and lead to a step change in performance of Matta's AI copilots across all manufacturing. The project is a collaboration between Matta and King's College London. Matta is a start-up company spun out from the University of Cambridge that develops artificial intelligence (AI) copilots for advanced manufacturing processes. These copilots can monitor, control, and optimise production to achieve higher quality, lower waste, and faster innovation. Dr Alhnan at King's College London is a leader in digital tablet fabrication and pharmaceutics and holds four granted patents in the field.