Research spotlight: Dr Caitriona Ryan
(Maynooth, October 21st 2020) Machine learning and artificial intelligence have not yet been exploited to a great extent in manufacturing. I-Form aims to develop new statistical technologies to solve key research questions in advanced manufacturing. Dr Caitríona Ryan is a Postdoctoral Researcher at NUIM. She specialises in Materials Processing Feedback and Control. Dr Ryan is a statistician and is excited about applying her knowledge and skills to a broad range of advanced manufacturing and engineering applications, as part of I-Form.
Detecting where something is going wrong, when a part is being additively manufactured is vital. Indeed, potential software tools could make a decision and perhaps pause or stop a build to fix problems and save resources. Dr Ryan has done important research in this area using statistical and machine learning techniques to detect anomalies in the data. Investigations into metal laser-based powder bed fusion (PBF-LB) 3D printing technology are underway, with the aim of detecting defects in real time. For example, data analytics using process monitoring data is key to identifying defects such as porosity in the built part. Images of the powder bed have also been used to detect larger scale defects such as elevated edges, as they become apparent. Thus engineers are alerted to these potentially serious defects during a build.
A number of questions will be addressed in the period ahead, including
- Will current approaches work for more complex structures?
- Can we find useful models to predict mechanical properties and porosity, given in situ process monitoring data?
- Can anomaly / potential porosity information, be fed back during a build to improve the build?
- Can we harness more information from these rich "big" datasets?