Data science for advanced manufacturing

Helping manufacturers use data to improve the efficiency and flexibility of processes, and apply new technologies with confidence

The decreasing cost of sensors, improvements in remote sensing technology, and the increasing ease of data collection has led to a revolution in industrial measurement. Manufacturing processes can now be monitored, maintained, and adjusted as they progress, improving product quality and reducing the number and duration of unplanned stoppages. Supply chain data can be used to inform product assembly and reduce scrap rates. To achieve these goals, industry needs reliable data analysis tools that can identify the significant features in a dataset that affect a process.

We are working with industrial and academic partners to address the challenges of industrial data analysis. Our work ranges from developing methods for smart assembly that minimise measurement cost, to time series analysis for preventative maintenance; and from uncertainty evaluation for sensor networks in factories, to development of digital twins for high-value manufacturing.

Our work involves the following themes:

• Trusted algorithms and software for decision-making under uncertainty for control, conformance assessment and product verification
• Techniques for uncertainty evaluation for advanced manufacturing measurement processes such as in-line and on-tool measurement, sensor networks and additive manufacturing
• Tools to address the data science needs of the automated factory of the future and digitally enabled supply chains, including data fusion, time series analysis for predictive maintenance and AI for process control
• Development of methodologies for development and updating of digital twins that ensure that uncertainties associated with the data stream and the model are handled correctly

Contact us