A Post-COVID Vision, by Dr Alan Hodgson

A monochrome camera.

A monochrome camera.

The time has now come for us to be considering the changes that the current pandemic will make to our industry. The purpose of this article is to look at the reshaping of one particular technology with the intention of prompting a similar appraisal of others. It is does not attempt to identify all the solutions but to at least propose a framework for the questions.

Applications within our industry, such as fingerprint and identity authentication, will undoubtedly be different as a result of the pandemic. However, it is also important for us to consider industries with less direct relevance, in order to unearth opportunities and threats. The aim of this article is to use the machine vision industry as an example and consider some of the potential implications for tax stamps and packaging.

Machine vision

Machine vision has become an important technology in a number of industries and we should consider the implications for product authentication. In essence, machine vision systems acquire an image of some sort and make decisions on the basis of this. It is that capability of automated decision making that makes it attractive to many industries.

Machine vision is not a new technology to us, with good examples already in existence within identity authentication such as e-gates and passport readers. Production monitoring and track and trace systems from companies such as SICPA, OpSec Security and Lake Image Systems utilise machine vision on production lines to read product markings. Integrated into wider production monitoring solutions, these vision systems form part of track and trace, auditing and business intelligence platforms.

Machine vision forms part of a wider authentication platform within the tax stamp application space. In combination with machine readable unique identification marks such as QR codes, they enable the monitoring and tracking of individual items throughout the process. But machine vision also has the potential to go much further, with the checking of features such as holograms on individual tax stamps, and even to validate that the tax stamp has actually been applied.

It is also a technology that works behind the scenes in many application areas. For the currency industry, one hidden application area for machine vision is in banknote checking. Companies like Weihai Hualing Opto-Electronics in China have image sensors specifically for banking applications.

Machine vision technology is evolving rapidly, and this may raise new opportunities and threats for our industry. A few years ago, the key application areas were in transportation, Industry 4.0 (the Smart Factory concept) and Internet of Things (IoT). However, the pandemic has resulted in a refocus of this that may have significant implications for tax stamps and product authentication.

Machine vision refocused

The current pandemic resulted in some immediate problems for the machine vision sector. Key markets such as automotive and assembly line building suffered substantial reduction. However, the machine vision industry refocused on sectors showing substantial growth during this period and I would like to highlight three of these as being of particular relevance.

The first is the manufacture of consumer electronics and the supporting semiconductor and electronics sectors, where machine vision is used extensively. These have done comparatively well throughout the pandemic as consumers have spent more time and money on online activities and purchasing. We should be factoring into our plans further advances in smartphone and consumer electronics technology, as these have remained well funded throughout the pandemic.

The second is the use of machine vision in the field of logistics, particularly in the fulfilment of e-commerce. Machine vision is being used to recognise and select products throughout the supply chain. Given that consumers in particular have become more accustomed to e-commerce, this is an area where we should perhaps give some consideration, particularly around the changing needs for product recognition and authentication in this sector.

Finally, and not surprisingly considering the pandemic, machine vision has found substantial opportunities in life sciences. From medical X-ray interpretation to COVID-19 bioassays, this is a substantial area of interest. One key observation for us is the strong linkage to other technologies, notably artificial intelligence (AI).

A technology ecosystem

Like many of the technologies recently discussed as insurgency threats (TSTN September 2020), we should not be considering machine vision in isolation. It exists in an environment where other complementary technologies are evolving in parallel and that has not changed in the post-COVID world. One key complement we would be wise to consider is the implication of the development and deployment of the combination of machine vision and AI.

As a particular example, AI is gaining traction in healthcare machine vision and this is a trend we should remain aware of. As part of the widespread COVID response, flexible deep learning models are being deployed that effectively allow machine vision to learn how to recognise and classify objects by their image. While this is currently being developed in a life sciences/healthcare setting, we should consider if the same learning algorithms could be deployed for product authentication in supply chains, as well as consider their effect on future tax stamp implementations.

Considering opportunities and threats

In TSTN June 2020, Juan Carlos Yañez, Chairman of The International Tax Stamp Association (ITSA), advised that: ‘new technologies bring new possibilities, but they should be considered suitable replacements of tax stamp best practices only if they can prove that they are more effective, less expensive or preferably both’. We should use that guidance when considering the potential impacts of technologies such as machine vision.

Here are a couple of examples in a machine vision context but I do suggest you consider other post-COVID changes in a similar fashion.

There looks to be no let-up on smartphone development through the pandemic. AI technology is being integrated into the current generation of smartphones, giving them enhanced machine vision capabilities. Given that these smartphones are going into the hands of the population at no cost to our industry, we should consider this development further; it passes the ‘Juan Carlos test’ above.

This combination of machine vision and AI in widespread deployment raises many additional questions. What features on a package or tax stamp will be most recognisable? What features would be made indispensable or redundant through this? How could a technology insurgent deploy these in collaboration or competition to us?

In summary

There is a need to consider how the post-COVID world will differ from the past to identify the opportunities and negate the threats. There is a role here for associations such as ITSA and the International Hologram Manufacturers Association, together with publications such as Tax Stamp & Traceability News™ and the post-COVID conference agenda. I suggest that there is a need to take this debate further.

This article considers a technology pertinent to the physical/digital interface for tax stamps. This interface between the physical and digital is the province of the Digital Document Security™ conference, which, in the mid-COVID world, will be held online from 30-31 March 2021. This would be a good place to explore these issues for tax stamp programmes.

The recently published report ‘Tax Stamps & Traceability: A Market Analysis and Technical Update’ provides many examples of the use of machine vision in tax stamp programmes. In particular it covers a number of case studies, over a wide geographic spread, describing implementation of machine vision in the wider context of revenue enhancing processes.

For those looking for more detail on machine vision it can be found elsewhere (ie. in TSTN’s sister publication Authentication News, July 2017). This was the start of a series of articles on ‘Smartphones in Authentication,’ as introduced in TSTN the same month. The whole topic of smartphones in authentication will be the subject of a Reconnaissance International publication of the same name in 2021, so look out for this.