Saritha Unnikrishnan

Current PhD Student

Predictive Modelling of Droplet Characteristics using Machine Learning for Emulsion Process Optimisation

In pharmaceutical industries, the quality evaluation of healthcare products is typically based on the examination of samples under the microscope by analysts. There exists a high risk of subjectivity and lack of precision in quality evaluation in industries where the process is entirely based on human examination. Recent advances in industrial automation has imposed tremendous competitive pressure on chemical industries to rapidly improve the techniques applied for product quality analysis and process control. The objective of the current study is to develop a machine learning based predictive model for the automated quality evaluation of oil in water emulsion products. The oil droplet characteristics of the emulsion product are extracted through automated image processing of microscopic images. These characteristics form a multivariate data set which is further used to train machine learning models to predict the quality of the product and the optimum process time required to achieve the desired quality. The model will be validated and integrated, for future work, with real-time image acquisition to classify in-line product quality and to predict the optimum process time.

Presentations

• Emulsion quality evaluation using automated image analysis, S. Unnikrishnan, J. Donovan, R. Macpherson, D. Tormey, Proceedings of International Conference of Innovative Design & Manufacturing (ICIDM), 2017, (Accepted for publication, in-press).
• Multi-Response Optimisation of Image Processing Parameters Using Central Composite Design and Desirability Function, S. Unnikrishnan, J. Donovan, R. Macpherson, D. Tormey, Irish Manufacturing Conference (IMC), 2017, Sligo, Ireland.
• Automated Analysis of Micrographs for Emulsion Quality Evaluation, S. Unnikrishnan, J. Donovan, R. Macpherson, D. Tormey Microscopy Society of Ireland Annual Symposium (MSI), 2018, Sligo, Ireland (Best Oral Presentation Award).
• A Comparative Statistical Study of Micrograph Classification using Repeatability & Reproducibility Analysis, S. Unnikrishnan, J. Donovan, R. Macpherson, D. Tormey, Conference on Applied Statistics, 2018, Galway (Poster Presentation).
• Machine Learning for Automated Micrograph Classification, S. Unnikrishnan, J. Donovan, R. Macpherson, D. Tormey, Irish Medtech Conference, 2018, Galway (Poster presentation)

Collaborations

GlaxoSmithKline, Sligo, Ireland

School/DepartmentProgrammeSRC/RRG
Engineering and Design PhD PEM

Funding

President's Bursary

Supervisor(s)

Dr. David Tormey. Dr. John Donovan.