'工学部/工学系研究科 トピックス' blog (64135253074) of hub id 20511701

TOPICS

Anicia Zeberli (D3 : at the time), Department of Chemical System Engineering, received Research Encouragement Prize of the Division of SIS at SCEJ 86th Annual Meeting

 

On 31th March 2021, Anicia Zeberli(D3 : at the time), Department of Chemical System Engineering, received Research Encouragement Prize of the Division of SIS at SCEJ 86th Annual Meeting.

 


〈Name of award and short explanation about the award〉
"Research Encouragement Prize of the Division of SIS at SCEJ 86th Annual Meeting" This award is presented to individual young speakers of general lectures who are expected to contribute significantly to the future development of the fields of science and technology and industry, and who have made oral presentations worthy of the award.

About awarded research〉
The introduction of the modern tools of Industry 4.0 in terms of digitalization and the incorporation of computational techniques from process systems engineering offer valuable opportunities within the pharmaceutical industry. In my work, I aim to demonstrate possible applications within biopharmaceutical production, in addition to identifying and tackling potential obstacles to their implementation.
The focus of the performed case study lays on changeover processes, with are the processes enabling the production of several products on the same line by adapting the production line to new products or product sizes and ensuring an aseptic production environment. Changeover processes occupy a significant portion of the drug product manufacturing operations and are product independent which allows the direct application of insights to other processes. The high number of steps within this process makes it possible to apply the failure detection and diagnosis approach within an active run. Statistical feature monitoring was applied to overcome the challenges related to the nature of pharmaceutical data, e.g.,  of the varying length of steps. For classification, the machine learning algorithm Random Forest and a control level approach were compared, where comparable results were obtained. The approach allows diagnosing both the source and the location of the problem. This combined prognostic and diagnosis application enhances the online operation support and makes immediate maintenance action possible, leading to improved productivity and efficiency.

〈Your impression & future plan〉
Without the continuous support of Prof. Sugiyama and the members of the Hirao-Sugiyama laboratory it would not have been possible. I am excited to see, where the research journey takes us in the future.