Virtual sensors for Insect production

In a collaboration between the Danish Technological Institute (Teknologisk Institut) and Consibio, two projects have been initiated – a Bachelor project and a Research & Development project. This project description refers to the Research & Development project. Johan Egsgaard Thomsen, co-founder of Consibio, will during this semester write a R&D-project based on the subject of virtual sensors in bioproduction processes.

A virtual sensor continuously estimates the state of one or more process parameters, which cannot be measured directly, based on other easily measurable parameters (like temperature, pH, gaseous CO2 etc.) and mathematical models, which are solved for in real-time. Furthermore, virtual sensors allow physical sensors that are separated over large distances to be used in computing an estimate, if implemented on distributed systems such as cloud computing platforms. This feature can be very practical when a virtual sensor is used in a large production facility.

Biological processes are due to their inherent complexity and variability, notoriously difficult to model and predict. This entails the need for continuous measurements of the process. However, physical sensors are often expensive, unreliable or unpractical to utilize due to many different factors such as restrained place, demands for sterile conditions, or simply due to the fact that the desired property cannot be measured by an on-line sensor. The concept of virtual sensors poses a viable solution to circumvent such limitations.

Virtual sensors can be applied to many different biological processes, but the production of insects for protein production will be the main case in this project. The Monesco platform will utilized for the monitoring and controlling of the bioprocesses, and the following physical sensors will be used to implement the foundation of the virtual sensors: