In the SmartFD project we aim at using mathematical modelling to improve the development of protein drug (biopharmaceuticals) manufacturing cultivation process for optimal solutions at reduced costs and labour. The focus of SmartFD will in particular be the creation of intelligent model-based tools to develop the culture media and the feed strategies to be used in manufacturing processes of fed-batch and perfusion. The market of feeds and media was nearly $3.2 billions world-wide in 2014. Performing culture media and feeds are keys for the success of mammalian cell-based cultivation process to obtain high yield and, as importantly, the desired quality of the produced biopharmaceutical. Such tools will bring a significant advantage both for the medium/feed manufacturer and for the industry developing new culture processes involved in the project. Furthermore, the project will respond to two presently unmet needs: the absence of methods to develop feeds/media supporting perfusion culture, and the possibility to use mathematical modelling to simulate, predict, optimize and control a given quality attribute in continuous culture system. By nature, such continuous process enables tuning a given quality attribute in a dynamic and reversible way, appealing for design based on mathematical modelling. Finally, an important aspect is that the rational approach SmartFD is taking, based on process modelling, will generate a process more robust compared to today’s method of trial-anderror , strongly supporting a Quality-by-Design approach, asked for by the manufacturers and the Health Authorities.

Project team

The project is a collaborative effort between 3 groups at KTH, and industrial partners GE Healthcare and Cobra Biologics.

Division of Decision and Control Systems KTH

Cell Technology Group KTH

Division of Glycoscience KTH

Project funding and duration

This is project is funded by VINNOVA, Sweden’s innovation agency, VINNOVA, with duration September 2017-September 2019.

Håkan Hjalmarsson
Professor of Signal Processing

My research interests cover system identification, process modeling and control, and communication network