The smart sensors enhancing safety

From the moment that chemicals arrive at factories until an end product is bottled, those on the manufacturing line must diligently follow rigorous cleaning and sanitation procedures on a regular basis to eliminate any chance of microbial contamination. This can lead to undesirable changes in the product, which may lead to wastage and product recalls, as well as losing the trust of consumers. Therefore, manufacturers carry out microbial analysis at every step of the manufacturing process, from the arrival of chemicals to the bottling of the product. This is a vital but time-consuming and resource-intensive process. 

As part of the digitisation of its supply chain, global consumer goods company P&G (Procter & Gamble) is using advanced technologies and the Internet of Things (IoT) to make its manufacturing more efficient. As the producer of household names such as Ariel, Oral B, Head & Shoulders, and Fairy, products must meet strict safety regulations and quality checks.

P&G’s processes involve regular cleaning and thermal sanitisation of the pipework. Establishing a truly clean state before sanitisation is challenging without the aid of monitoring devices. The company’s testing for microbial contamination is typically carried out ‘off-line’ or ‘at-line’, where samples are analysed away from the production line – in a separate lab for example – in a process that can take up to three days. The company recognised an opportunity to enhance these procedures in the production of its Fairy washing up liquid in a London-based factory. It actively sought a sensor technology that could be used ‘in-line’ to directly monitor the inside of the pipes for microbial contamination in real time. By implementing such sensors, P&G technicians aimed to gain accurate information on the product and pipe conditions, focusing the need for external testing. This would allow them to make informed decisions on the cleaning intervals within the factory, reducing manual effort. 

Sensing in industrial environments

Sensors and instrumentation play a central role in the transition to Industry 4.0, with increasing use across many industries, including food, agriculture, chemical, and defence, to collect real-time data for monitoring and controlling processes. For example, the data collected in processes such as temperature regulation, pressure control, chemical composition, or bacterial growth can inform quality control, predict maintenance needs, and improve processes. Better control of these processes is also crucial for waste management.

The sensor P&G required had to be reliable, contactless and ultra-sensitive. It also had to be durable enough to withstand the chemicals and pressures in industrial pipes over a long period, as well as low cost for large-scale implementation to be economically feasible. 

Many sensors in development are sensitive enough to detect small bacterial fluctuations, but this increased sensitivity makes them more vulnerable to external vibrations and temperature changes, which can interfere with microbial measurements. Vibrations can cause tiny displacements, or misalign optical or electronic components, within the sensor, leading to incorrect readings. This is a significant challenge in industrial settings where acoustic vibrations from pumps, fans and mechanical movements from seismic activity are unavoidable. For sensors that are particularly sensitive, even minute vibrations can introduce noise into readings and affect results. Slight changes in temperature can change a sensor’s sensitivity, as sensors are often calibrated to work at a specific temperature. Commercially available sensors that can withstand these environments are generally much less sensitive. After an extensive market search, the P&G team could not find a suitable device and realised that innovative engineering and advanced technology were needed. 

Turning to UK research

During the technology scouting process, P&G’s external Connect and Develop team became aware of work being carried out by a multidisciplinary team based at the University of York that specialised in innovative sensor technologies, led by Professors Thomas Krauss and Steve Johnson. P&G approached them to adapt their work for a pilot project funded by Innovate UK, and the team eagerly accepted the challenge. 

Through the project, the team discovered that some of the sensor technology they had been working on, originally developed for biomedical disease diagnostics, could potentially be repurposed to detect the bacterial biofilms that P&G wanted to target. This technology, a patented optical-sensing device called the chirped guided-mode resonance sensor (chirped-GMR) (see ‘Guided-mode resonance explained’), showed promise in early experiments. The next step was adapting it for industrial use and changing its design.

The research team varied the spacing of the grating elements along the length of the sensor. This is comparable to having a row of slits that gradually get closer together or further apart. This ‘chirped’ design – meaning that the frequency gradually increases or decreases, similar to a bird’s chirp – allowed them to represent the spectral signature of the resonance as imaging data. For example, a change in spectrum becomes a change in position on the sensor. 

Such a change can be readily imaged with a simple low-cost CMOS (complementary metal-oxide semiconductor) camera. In this way, the chirped GMR sensor combines the sensing and the spectrometer function in a single element.

To address the issue of mechanical and thermal noise that could interfere with industry measurements, the researchers changed this grating structure further into a bowtie design. The mirror symmetry of the bowtie helped to cancel out noise from physical disturbances by moving both resonances in opposite directions. When a physical disturbance, such as vibrations, occurs it typically affects both sides of the bowtie equally but in opposite directions, leading to cancellation of unwanted noise. 

This precise engineering ensured that the signal produced from the sensor was only caused by the change in refractive index. 

Real-world testing

Integrating a device developed in an academic setting into the industrial manufacturing process is challenging. The new technology needs to be as reliable or better than existing methods before manufacturers can complete the transition. 

The first phase of the project installed the new sensor technology in a controlled environment at P&G’s research labs in Reading. P&G’s microbiologists intentionally added bacterial spikes with independent experts testing whether the sensors could detect biofilms with the required sensitivity and thus provide early warnings of microbial contamination. The sensors consistently performed well in laboratory tests and were verified by P&G at the pilot plant scale, so now the project has moved to the manufacturing demonstration phase. 

According to Andy Dorset, Senior Director of Quality Innovation at P&G, “the manufacturing demonstration phase marks an exciting moment for using real-time in-line sensors to detect microbial contamination throughout the entire supply chain, from suppliers to export trucks. These sensors, connected to a digital platform, can provide insights into how the different parts of the process can potentially impact microbial contamination.”

P&G has installed multiple sensors to monitor the presence of microbes at different stages of the end-to-end Fairy liquid supply chain. The sensors have been implemented at an external supplier in Manchester, as well as throughout the product manufacturing process in London and within trucks that transport intermediate materials to a central European location, where the product is bottled.

Dorset notes the potential of these sensors to replace the current manual methods, describing these as “labour-intensive and sample-heavy”. He emphasises that “if successful, this sensor-based approach would provide information much earlier than manual methods. It would enable faster decision-making to address any microbial issues, and also lead to productivity gains by digitising manual and paperwork-intensive processes.”

It will take time to complete this phase of the project and the sensors must be able to alert the company to unplanned contamination on the production line to prove successful. P&G has to establish the effectiveness of the sensor-based approach compared to the existing manual microbial analysis before it can consider a complete replacement of the current process. 

Dorset further explains that the new approach could affect the organisation’s cleaning and sanitisation frequency. “The sensors would serve as in-line monitors, notifying us when we need to take action instead of following a fixed schedule.” He adds: “With these benefits in mind our intention is to expand the implementation of these sensors and digitisation to encompass our microbe-susceptible supply chains. This is a pioneering step forward in ensuring product quality in our supply chain.”

Beyond washing up

The York research group has founded a spinout company, Phorest Diagnostics (drawing inspiration from the word ‘photonics’ – the study of light), to continue developing its patented chirped-GMR technology. This will also allow them to draw on the challenges and learning experience encountered so far.

“Installing the sensor in an industrial setting has been a steep learning curve. We needed to consider not only the robustness of the technology, but also its interfacing with the process control software” explains Professor Thomas Krauss. “Moreover, all the fittings needed to meet Clean Design specifications, which is not something we worry about in the academic setting.”

Phorest Diagnostics aims to expand into other industrial sectors beyond washing detergents where in-line bacterial monitoring is essential, such as food production and transport.

“Looking ahead, we are very excited about the prospect of a fully digital supply chain, where process engineers can remotely monitor the quality of product end-to-end, from precursor to finished product,” Krauss adds. “To this end, we have already installed a sensor on the truck of a major logistics company.”

As industries continue to embrace smart manufacturing, innovations such as the chirped-GMR sensor will play a crucial role in enhancing efficiency, reliability, and overall productivity.