• Cambridge Advanced Research and Education in Singapore (CARES), the University of Cambridge's first overseas research centre, together with the A*STAR Institute for Infocomm Research (A*STAR I²R) has developed a digital twin platform that uses artificial intelligence (AI) and real-time plant data to enhance fault detection, system monitoring, and predictive maintenance in pharmaceutical manufacturing.
• The technology will be commercialised through Chemical Data Intelligence, a CARES spin-off, and will be made available to the Pharma Innovation Programme Singapore (PIPS) Consortium.

-- Pharmaceutical manufacturing demands consistency, high precision, and continuous improvement to ensure the safety and quality of medicines. The new platform automates the development of digital twins, which are virtual replicas of production lines, to optimise plant operations, detect anomalies early, and support data-driven decision-making.  

The technology has shown key advantages of an ontology-based digital twin to improve system understanding, refine predictions, and supports the addition of flexible, new workflow agents for future digital development and manufacturing.

CARES led the development of the digital twin's ontology and physical modelling, while A*STAR I²R developed an AI agent for anomaly detection. This AI agent combines physical models, which utilises the knowledge from the ontology, with the data-driven models trained on sensory data such as temperature, flow rates, and pressure, to detect issues like mismatched flow rates or abnormal tank levels before they escalate.

The system was demonstrated on a real-time manufacturing testbed provided by another CARES spin-off, Accelerated Materials. Hosted on Microsoft Azure cloud, the platform enables researchers and engineers to monitor plant performance in real-time, simulate production scenarios, and test responses to potential faults.

Prof Alexei Lapkin, lead Principal Investigator on the project, reflects on future developments, "the project has multiple outputs; one is the technology for rapid development of digital twins based on the ontology of first principles models. This technology is now being turned into a commercial product by another CARES spin-off, Chemical Data Intelligence, to make it available to the pharmaceutical companies under the Pharma Innovation Programme Singapore (PIPS) Consortium"

Dr Lianlian Jiang, Co-Lead Principal Investigator of the project, Unit Lead of Digital & Sustainable Manufacturing at  A*STAR I²R, said: "The AI agent in this ontology-based digital twin can be extended beyond anomaly detection to support quality monitoring, production scheduling, and resource planning. By embedding domain knowledge into the system, the technology helps capture and transfer critical expertise while complementing staff expertise. We are glad to work with CARES and industry partners to further scale this technology to meet the needs of more complex manufacturing environments."

Watch a video explaining the technology of the project: https://youtu.be/A2VZNdIiXEs

The project, From Digital Twins to Real Time AI-Supported Plant Operations, is supported by A*STAR under its Pharma Innovation Programme Singapore (PIPS) (Award no: M23B3a0015). The Agency for Science, Technology and Research (A*STAR) is Singapore's lead public sector research and development (R&D) agency.