The Sensor is an advanced process monitoring system that can analyze the signal of the measuring device and generate new information of the machining state through the popular mathematical algorithm. Sensor for the calculation of the signal is mainly based on two models, namely, model-driven and data-driven. Sensors are widely used in industrial production, the purpose is to improve and promote the processing of the production, product quality and safety. But the Sensor has not been applied to the biotechnology industry. In the field of bioprocessing process monitoring and control, the system is required to have the ability to measure all process variables. In addition, there is a high demand for system continuity and real-time. The more complete the measurement, the higher the reproducibility and efficiency of the process, the better the product is produced. The analysis of food and pharmaceutical processes in the United States is closely related to the analysis and control of biopharmaceuticals. The goal is to obtain a suitable process, by measuring the process variables to ensure product quality. And this process is monitored by an efficient and suitable sensor system. This paper reviews the existing and emerging Sensor applications and the difficulties that need to be overcome in the biotechnology industry.
The development of Sensors requires the need for industrial production, which can guide the design of Sensors. Table 1 shows the online monitoring points that may be involved in large-scale bio-processing. Mammalian and microbiological bioprocesses are different. Although there are many differences between processes, monitoring requirements are similar in many ways. Choosing the most reasonable analytical technique requires careful design. Another important requirement for all biological processes is to monitor information that should generate transient process status on-line. Key processes can track the parameters of the upstream process to better adjust to achieve more efficient control, optimize harvest time and improve the quality of the final product.
Various analytical techniques have the potential to address the mapping's priority requirements. This has important implications for signal processing, estimation and prediction. Near infrared and mid - infrared spectroscopy, in situ microscopy, 2D fluorescence spectroscopy and high performance liquid chromatography were added to the possibility of new specific ligand in the biomarker target in the culture medium. This includes aptamers and short peptide recognition elements that are sufficiently stable to withstand the harsh conditions in the bioreactor. The method based on microfluids and nanoparticles provides new possibilities for the analysis of the availability of Sensors.
In situ NIR / MIR spectra with multivariate data analysis meet many desirable ideal Sensors: samples need not be removed, data is generated in real time, without additional reagents, easy to install and sterilize in situ probes, potentially While measuring several components to facilitate multiple use and low cost operation. The initial investment cost is high, the spectral data are cluttered, and the sensitivity of the compound is determined to be too low. Spectral analytes may overlap, the background deviation of the culture medium is predicted, and validation is difficult.
With the imaging software, the ISM probe can monitor the important parameters of the suspension cells in the bioreactor in real time and can be used in various culture systems. In addition, the ISM probe can be sterilized in situ, allowing the probe to continue to operate in the reactor. The disadvantage is that bubbles and cell debris may interfere with cell image analysis and parameter evaluation.
Regardless of the advantages or disadvantages of the Sensor, it is still not actually applied. Because the application of new technology requires a long period of repeated verification. These certifications require professional knowledge and a lot of capital investment, if the benefits of Sensors is not ideal, put into use will have a huge risk.
In order to service these obstacles as soon as possible, Figure 4 lists some key guidelines. First of all, need to be clear what is the purpose of using this technology, want to get what kind of income. Second, the Sensor has different algorithms and computational models, lists different techniques, compares their different metrics, and compares their expected goal fulfillment. Third, select the appropriate technology, conduct a feasibility study, and build SOP. Fourth, the establishment of the technical operating environment, including the algorithm and the reactor connection, sampling devices, the use of methods and so on. Fifth, performance evaluation. Sixth, the final validation in normal production.