Abstract
Among the greatest challenges in soft sensor development for bioprocesses are variable process lengths, multiple process phases, and erroneous model inputs due to sensor faults. This review article describes these three challenges and critically discusses the corresponding solution approaches from a data scientist’s perspective. This main part of the article is preceded by an overview of the status quo in the development and application of soft sensors. The scope of this article is mainly the upstream part of bioprocesses, although the solution approaches are in most cases also applicable to the downstream part. Variable process lengths are accounted for by data synchronization techniques such as indicator variables, curve registration, and dynamic time warping. Multiple process phases are partitioned by trajectory or correlation-based phase detection, enabling phase-adaptive modeling. Sensor faults are detected by symptom signals, pattern recognition, or by changing contributions of the corresponding sensor to a process model. According to the current state of the literature, tolerance to sensor faults remains the greatest challenge in soft sensor development, especially in the presence of variable process lengths and multiple process phases.
Highlights
The biologization of the manufacturing industry is leading to more and more processes that were previously based on chemical synthesis being replaced by biotechnological processes (Buyel et al, 2017)
Based on an overview of the status quo of soft sensor development and online implementation, this review article describes the challenges of variable process lengths, multiple phases, and sensor faults, and critically discusses the corresponding solution approaches
The challenges are considered both individually and synchronously, and the solution approaches are evaluated in terms of their practicality and applicability to bioprocesses
Summary
The biologization of the manufacturing industry is leading to more and more processes that were previously based on chemical synthesis being replaced by biotechnological processes (Buyel et al, 2017). The digitalization of these processes is leading to more transparent, lowerrisk, and more efficient biological manufacturing (Scheper et al, 2021). At the intersection of these two trends—biologization and digitalization—a multitude of new technologies and approaches have emerged in recent decades. These include, in particular, advances in the fields of data science as well as monitoring and control technology for bioprocesses (Steinwandter et al, 2019). Despite advances in bioprocess monitoring, many relevant process variables are still determined offline using laboratory analyses. On this basis, a prediction is made about the expected future behavior of the process. The development of soft sensors is a remedy to this situation
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
