Digitalisation is a highly desirable activity for a majority of businesses wishing to remain abreast of the latest technologies that can be used to optimise processes. Daniel Povey, Higher Research Scientist at NPL discusses the traceability of data produced by a digitalisation solution.

 

Introduction to Digitalisation

Digitalisation is a term often used in reference to the use of digital technologies to update, transform, or otherwise generate additional value from processes and/ or opportunities. In industry, and particularly in manufacturing, it is most commonly used to refer to the implementation of systems that capture information about a process, which can then be used in analysis and optimisation. Broadly speaking, it is a desirable activity for most businesses to undertake: the prospects of improved process insight, efficiency, and performance are hard to ignore.

Digitalisation could refer to the implementation of any number of modern digital technologies. From Internet of Things (IoT) sensors to robotic automation, or machine-learning-generated process insights. In all cases, however, digitalisation is about the generation of data and using it to inform decisions.

Confidence in Data

One of the most important aspects to understand when using data to inform decision-making is the need for confidence in that data. Decisions informed by poor data are likely to be poor decisions, leading to disruptions, inefficiencies, loss of income, and at worst: risk to the health and safety of operators. To have confidence in a data-based decision, there must be a chain of traceability that leads from baseline measurements through transfer, processing, and analysis of data, up until the moment a decision is made.

Often the first step to establishing this traceability is to ensure that implemented sensors are fit for purpose. These measurement devices are the foundation on which the decision-making infrastructure is built, and any uncertainty introduced at this stage will propagate through to the end decision. Whether that decision is to implement a process alteration, estimate a production quantity, or implement a countermeasure such as a full-stop in the event of a major fault, it is vital to understand how that uncertainty contributes to the final decision.

Thus, it is critical that the sensors have an appropriate combination of accuracy, reliability, and precision. By using calibrated and validated sensors, demonstrating traceability back to national standards, confidence in the data is improved and the first link in the traceability chain is established. Confidence must be maintained throughout the lifecycle of the data, so any intermediate operations between data capture and decision-making must also be validated and verified. This will ensure that any uncertainty in the measurements is correctly identified, and its propagation tracked throughout the chain.

It is not always a simple task to ensure confidence in data: there are a huge number of variables that can affect a particular sensor’s viability for a specific purpose, as well as logistical considerations for the timely transport and analysis of data that could prove just as important as the measurements themselves. As such, the digitalisation of a process requires not only an intimate knowledge of the process that is being digitalised, but also the solutions being used to enable it.

Digitalisation Test Beds at NPL

In a collaboration with the University of Birmingham and Kings College London, funded by UKRI, the National Physical Laboratory (NPL) investigated the application of digitalisation technologies to face mask testing. Specifically, the possibility of using a combination of dimensional analysis and digital modelling to assess the performance of bespoke face masks in physical testing was explored.

The Tests

The key performance metrics of the face masks under consideration are the filtration efficiency of the mask filters and the leakage of the masks around the perimeter of the seal when donned to a human face. The physical tests themselves were based on those found in Standards, but adapted for improved reproducibility; modifications such as the use of a silicone headform in place of live human test subjects were implemented. A schematic of the test apparatus can be seen in Figure 1.

Figure 1: Schematic of the filtration efficiency test used in the investigation.

By fitting a face mask to the silicone headform, the test apparatus was used to measure the particle leakage and (using only the filter rather than the headform) the particle filtration efficiency of the face masks and their filter material, respectively.

Dimensional Validation

The fit of the mask itself was assessed using a contactless laser line scanner to identify gaps. Data from these dimensional measurements were compared against CAD models of the mask and head form to identify gaps in the fit, as well as deviations of the physical components from the models used in their manufacture.

Figure 2: "Heat-map" of the deviation of the fit of the physical mask and headform. 

 

Digitalising the Tests

The digitalisation of this test method takes the form of a finite element analysis (FEA) model, in which the silicone head form and the face mask assembly are modelled for contact analysis. It is hoped that, by validating this model using data from the physical testbed, a robust prediction of the performance of each head form and face mask combination can be produced. More specifically: filtration efficiency and leakage results from the physical tests are to be analysed with respect to the dimensional measurements performed on the tested headform/face mask combination. Correlations arising from this analysis are to be used to validate the virtual model – such that (with sufficient data used to inform the model) it would be possible to accurately predict the leakage and filtration for any given combination of headform and face mask.

Figure 3: Contact pressure modelling of the silicone headform and face mask.

It is hoped that the output of this investigation will lead to improved face mask test methods, as well as to enable the viable production of bespoke face masks – tailored to specific headform or facial scan data.

Conclusions

Digitalisation is a highly desirable activity for a majority of businesses wishing to remain abreast of the latest technologies that can be used to optimise processes. In all cases, but particularly in manufacturing, the traceability of data produced by a digitalisation solution must be completely traceable in order to minimise risk to employees and consumers. This traceability requires an overlap of highly specialised skills and subject knowledge in both the process and its digitalisation.

As the UK’s National Measurement Institute, NPL’s role is to support UK industry – meaning that the digitalisation of industrial processes is an area of active interest and research. At NPL, we are happy to discuss potential projects involving the development and implementation of digitalisation solutions.

Contact

Daniel Povey – Higher Research Scientist
Manufacturing Metrology group
e: This email address is being protected from spambots. You need JavaScript enabled to view it.
w: https://www.npl.co.uk/about-us/locations/north-england/

HOW CAN WE HELP YOU?