Manufacturing systems of all hues are becoming ever more increasingly complex because of the number and sophistication of the operations in the production chain. This is alongside all the various multiple structural characteristics. One approach is to exploit smart intelligent sensing.
The intelligent sensor is best described then as a sensor with built-in reasoning or information pre-processing that provides high-level knowledge. These types of sensor are distinct from microminiature sensors and embedded and distributed sensors. The microminiature sensors are integrated circuit (IC)-based with built-in signal processing.
These intelligent sensors are the fundamental building blocks of modern smart factories. They enable sensor-supported production resources such as machines and robots to configure, control, manage and optimize themselves. Precise, reliable sensor data is now more essential than ever.
The system of intelligent sensing exemplified here is part of Industry 4.0. or the Fourth Industrial Revolution which is the next step in digitalization of the manufacturing sector. It covers the rise of data gathering management connected to human-machine interfacing and developments in robotics. These new production systems have to generate reliable and contextual information from a wealth of data and make it available to people, machines and IT services.
The benefits of intelligent sensing in the life sciences are many fold:-
- Tighter control of the process in terms of more accurate monitoring.
- Improved hygiene status
- Better control of fermentation process especially in continuous operations where high value cellular fermentations are operated as in probiotics production, cell cultures for a range of secondary metabolites. That also implies a better understanding of the process that leads to improvements.
Intelligent Sensing and Filtration
Filtration is a significant downstream processing operation in both food and beverage, and in life science applications generally. It has not been without its teething issues in terms of the detection of situations that arise in filtration as well as in maintaining performance and continued success of operation.
The most common problem in filtration is clogging and blockage of the filters. This shows itself as a drop in the permeate or filtration rate across the membrane and a steady rise in transmembrane pressure. If allowed to continue there is a complete end to filtration and with it lost performance, downtime in cleaning the membrane and even replacement of the blocked membrane simply through damage.
Strangely, detection of the loading status on a filter membrane is not as easy to achieve as expected. We’ve already identified complete clogging as an easily recognizable cause of pressure drop in the filtration system. However, if it is feasible to conduct online monitoring of the actual loading status on the membrane then it should be possible through electronic analysis to start making significant in roads into process development. We’ve already alluded to this by discussing some benefits but the subject is still in its infancy and there is significant work to show cost-benefit analysis and convince users that they can understand its application.
Many intelligent sensing systems need to be bespoke applications. A general sensing system needs to be designed for a plethora of processing situations where the filtration medium varies between types of fermentation as well during the change occurring during through operation. There are significant differences in particulate content and the abrasiveness of that filtration medium.
Generally, it is expected that the sensors detecting clogging for example are built into the membrane. That would then allow the integration of the electrodes which are part of the membrane into disposable cartridges. At the moment they are not cost effective.
When it comes to filtration the industries that would benefit from closer control are dairy processing, the wine and beer industry, meat and fish hydrolysate production. All these industries suffer with particulate blockage. These can take the form of physically large particulates such as muscle fragments and cells through to colloidal particulates as in hazes that afflict wine processing.
Applications
In the early years of filtration-related intelligent sensing, the Fraunhofer Research Institution for Modular Solid State Technologies at the EMFT in Munich, Germany reported on an intelligent sensor-filter (Alberti et al., 2012). This sensor was based on two micro-perforated sensor-foils with interdigital capacitors (IDC) and a filter membrane placed in between them. It was possible to monitor the change in permittivity the fluid upstream on the retentate side using that sensor placed on that side. The sensor on the permeate side of the filter then compensated for temperature and conductivity variation of the permeate itself.
An intelligent filter system has subsequently been developed by Wolftechnik Filtersysteme GmbH & Co. KG in partnership with Mittelstand 4.0-Kompetenzzentrum Stuttgart. The filter monitors a host of operating parameters which means that it can detect issues at the earliest opportunity. The benefit as with all a priori knowledge is that scheduling of maintenance and servicing can be organised in response to changes as opposed to having scheduled maintenance when it’s not needed. The other benefit is that the number of unscheduled disruptions is minimised as is the cost of maintaining an inventory for filter cartridges.
The interface is a screen which acts as a dashboard allowing the user to manage the whole filtration process via this single interface. The parameters are configured, monitored and all the data backed up and restored. The dashboard also means access to documentation like user manuals, specifications etc. The core system is a smart controller collecting and processing all the relevant parameter data for the filter system’s pressure vessels.
This smart filter is just such an example of the direction that control of production systems are taking. The Wolftechnik smart filter measures temperature, flow rate as volumetric flow rate and pressure. The system maintains a log file of all data recorded and transmitted. The operator’s IT system stores information about filters specifically in terms of spare parts and replacement filters. The digital maintenance and inspection schedules are created and operated on. Operators have the option of storing user instructions, servicing, maintenance and filter changes – especially useful in clinical management. The operator can also call up texts and images to guide them through the procedure. Such a set-up helps with correct installation but also makes sure maintenance and service tasks including filter changes are performed quickly and safely.
Real-time availability of data means a new approach to filter system control and organisation. You can easily exchange operating data with external service technicians. Data collected is used to analyse and diagnose faults quickly and rectify issues without delay. As well as automated spare part ordering and notification, there is also preventive maintenance and optimised production. Such a system is integral to management of HACCP in the food industry.
For The Future
One aspect for the future is whether sensory testing sensors might be employed to control future filtration processes. There are systems such as the electronic nose which uses sensor arrays and pattern recognition methods for checking whether products are adulterated or meet a particular product criterion. It should be feasible to use the same technology to check too on food safety such as pathogens or undesirable products in a filtrate. There may be possibilities too with nuclear magnetic resonance (NMR) metabolomics profiling as a means of immediate electronic sensory assessment (Steinsholm et al., 2020).
For example, the quality of permeates from industries such as dairy might be suitable. Different permeates from whey and cheese processing might be controlled at the filtration stage. Likewise, coffee and tea processing, juice quality, beer filtration, etc. Microfiltration removes suspended particles, emulsified fat and microorganisms to produce a light coloured permeate. Further processing using nanofiltration could be used to retain peptides whilst the ions and small organic molecules pass through. Nanofiltration also helps with dewatering thereby reducing further downstream processing costs that are then involved in drying. It might be feasible too with intelligent sensing to control via membrane filtration, the addition of water to help improve process yield or purity. This is part of the diafiltration processing and whilst it tends to add to costs, it could be cost effective if it significantly improves on both purity and yield further in the downstream process.
The quality of waste streams too might be open to this type of technology. Protein hydrolysates from meat and fish processing, rely on enzymatic hydrolysis (Aspevik et al., 2017). In the current processes, the issues of bitterness, rancidity are associated with these types of product streams which could be removed in one stroke if the separation process was either halted immediately or controlled to the extent where further processing could be controlled. Blue sky thinking of course, but you don’t have to wait for a human sensory evaluation. The same principle is feasible with the production of bioactive compounds such as antimicrobial peptides which are released by the same industry.
References
Alam, G., Ihsanullah, I., Naushad, M., & Sillanpää, M. (2022). Applications of artificial intelligence in water treatment for optimization and automation of adsorption processes: Recent advances and prospects. Chemical Engineering Journal, 427, 130011.
Alberti, M., Meixner, L., Rückerl, A., Eder, M., Endres, H.-E., Bock, K. (2012) “Sensor-Filter” – Intelligent microfilter system in foil technology. Procedia Engineering 47 pp. 212-215
, , , , , , & (2017). Valorization of proteins from co- and by-products from the fish and meat industry. Topics in Current Chemistry, 375(3), 53 (Article)
, , , & (2019). Valorisation of protein hydrolysates from animal by-products: Perspectives on bitter taste and debittering methods: A review. International Journal of Food Science & Technology, 54(4), pp. 978– 986 (Article).
Jiang, C., Wang, G., Hein, R., Liu, N., Luo, X., & Davis, J. J. (2020). Antifouling strategies for selective in vitro and in vivo sensing. Chemical Reviews, 120(8), pp. 3852-3889.
Mendoza, E., Andramuño, J., Núñez, J., Benítez, I. (2021) Deliberative architecture for smart sensors in the filtering operation of a water purification plant. J. Phys.: Conf. Ser. 1730 012088 DOI 10.1088/1742-6596/1730/1/012088
Niu, C., Li, X., Dai, R., & Wang, Z. (2022). Artificial intelligence-incorporated membrane fouling prediction for membrane-based processes in the past 20 years: A critical review. Water Research, 118299.
Seymour, A. (2023) BRITA hails its ‘biggest innovation’ since founder launched water filter jug in 1970. Catering Insight (16th February 2023) Accessed 20th March 2023
Wang, K., Zhao, W., Lin, L., Wang, T., Wei, P., Ledesma-Amaro, R., Zhang, A.-H., Ji, X.-J. (2022) A robust soft sensor based on artificial neural network for monitoring microbial lipid fermentation processes using Yarrowia lipolytica. Biotechnol. Bioeng. (Article)
Leave a Reply