Industrial Product Authentication is a relatively recent phenomenon so we thought it could be useful with an overview of methods and their pros & cons.
Why "Industrial" Product Authentication?
Before we go through the methods, let's first explain why we bother to add the word Industrial in the headline.
For long, there have been a cadre of highly skilled experts being able to differentiate legitimate items from fakes. Almost like artisans or detectives. Sometimes relied on by detectives and legal authorities, actually.
Similar to art experts they use their expertise to identify the genuine handbags or watches from the counterfeit.
It's a fascinating process but slow, relying on experts with a razor-sharp expertise that takes long to acquire. Obviously it's only feasible for the few items where the value warrants such an effort and cost.
Nothing you use to tell genuine spark plugs or ball-bearings from fake ones. For more mundane products we need something faster and more cost efficient; industrial product authentication.
The Principles of Industrial Product Authentication
Product Authentication works through comparing an item in front of you with a benchmark.
The benchmarking methods vary though:
Artisanal product authentication compare with the knowledge and experience of the expert (and maybe materials testing (similar to CSI)
Comparing items in the market with items that have left the production line of the legitimate producer
Comparing items in the market with an AI rule (the most novel approach)
How benchmarks are established and authentication is made varies, though:
1. Adding An Identifier to Items
This is the most common method, but which kind of identifier that is used varies. Typically they are
Labels with holograms
Labels with special ink (fluorescent or similar)
Adding scratches or micro features
RFID or some other electronic transmitter
Pros of Adding Identifiers
Familiar technology
Cons of Adding Identifiers
Interferes with production and production lines
Adding steps in the production requires substantial production interruption for installation, causing installation costs and leading to production shortfall
One more thing that can go wrong during production
Requires holding and managing a stock of identifiers, and the risk of running out
It's actually not the product that is authenticated but the identifier
Consumer trust in the process depends on their trust in the identifier
Can the identifiers be copied?
No added trust by consumers in the authenticity of the product itself
May interfere with aesthetics and product design
Often require special equipment to do the authentication
May not be possible to do in the field, on the fly
May require trained staff to do the authentication
"What you see you can copy"
How would you know what an authentic hologram look like?
2. Scanning Print, Codes or Labels on Products or Packaging
This method uses computer vision with extremely advanced image analysis, creating a digital twin of each item, storing it securely in the cloud for later comparison with an image taken in the field.
Pros of Scanning Print, Codes or Labels on Products or Packaging
No interference with production or production lines
As no steps are added to the production line, installation can be done during scheduled maintenance breaks. All that is needed is planning where to put scanners, power and network connectivity
Therefore: Lower installation cost, no production shortfall, faster start
No interference with running production
Boosts trust in your brand, as it is the actual product or package that is authenticated, not in the supplier of the authentication method
Does not interfere with the design of product or packaging
Authentication can be done using a smartphone in ordinary conditions, by anyone with the app
The digital twin can be supplemented with additional data, adding value to the product, value that might be possible to monetise
Cons of Scanning Print, Codes or Labels on Products or Packaging
Requires cloud storage of images between production and authentication
Requires permission to scan in the production line and send it to the cloud
No visible indication of items being possible to authenticate
Read more about authenticating by scanning print, codes or labels on product or packaging:
3. Scanning Surfaces of Finished Items
Similar to method number 2, you can scan the actual surface of the item instead.
Pros of Scanning Surfaces of Finished Items
Not dependent on the availability of print, codes or labels to scan
May be possible to repeat throughout the lifetime of the item, not only at first purchase
May be possible, if so, to supplement with data on life events of the item, service, repairs, transactions... creating a digital certificate of provenance and life events to augment the long term value of items
No package needed for authentication. It's the item, nothing else
Cons of Scanning Surfaces of Finished Items
Requires special rigs to control lighting and position of items at initial scan as well as authentication scans
Therefore slower and less likely to be possible to do in line with the production flow
Due to slower speed and special equipment, less versatile than method 2. More suitable for high value products. Therefore closer to the artisanal method mentioned first, but still much faster and with no need for the "detectives" experience
We have heard of some company trying to do something similar but scanning cardboard boxes, creating a "fingerprint" of print and surface patterns. We're doubtful about the feasibility of this method though because:
Sensitive for the impact of transport of packages
Once the package is gone, authentication becomes impossible
Still slower than methods 1 & 2 but without the advantage of repeat authentication and life events of method 3, using the surface of the product itself
Read more about counterfeiting of prestige brands:
4. Authenticating Against a Benchmark Rule Created With AI
Authentication in the field is made just like in method 2, but the comparison isn't made with scans of actual products sold in the market. Instead, samples of packages are scanned repeatedly and the images analysed with AI to create a benchmark rule to compare against.
Pros of Authenticating Against a Benchmark Rule Created With AI
No installation needed whatsoever. Just collect a sufficient number of samples, scan them and analyse the scans and let the AI model work
Faster roll-out across multiple production lines or factories
Makes product authentication possible also when scanners and uploading to the cloud are not available options
Lower costs than the previous three methods puts product authentication in reach for very mundane products, as in Fast Moving Consumer Goods, FMCG
Still as simple to authenticate in the field as for method 2.
Cons of Authenticating Against a Benchmark Rule Created With AI
The level of confidence depends on the design and looks of the packaging. Still, confidence levels even with as mundane products as shampoo bottles has been outstanding.
Read more about authenticating against a benchmark rule created with AI:
At Blue Cromos, We Provide All Methods Except the First, Because We Don't Want to Interfere with Carefully Optimised Production Lines
Scanning print, codes or labels on products or packaging is where we started, a method suitable for all sorts of products from spare parts to electronics or whisky.
Scanning surfaces of finished items we have added for prestige items, nothing can be allowed to interfere with design and where provenance and authenticity is key for long term value of items
Authenticating against a benchmark rule created with AI was developed to cater for FMCG, especially where in-factory scanning and uploading of data to our cloud is not possible or desired.
We Even Have a Fifth Method – for Components
All the four first methods are for finished products, but what about components, especially function-critical components? Counterfeiters have been able also to infiltrate supply chains. For that purpose we have developed a method where we scan the surface of incoming, unpackaged components to identify and authenticate them with scans made at the end of the production line of the sub-supplier.
Authenticating Components Opens New Possibilities
Using this method lifts quality assurance to new levels
In addition to the value of authenticating incoming function-critical components, they can be supplemented with data from the sub-supplier: "this item is intended to be assembled into product X, version Y". Data that you have originally provided to the sub-supplier and they have passed back to you as a "digital label" when they ship to you.
Your system can then double-check that you are indeed about to produce product X, version Y before any components of the current batch go into production. Moving quality assurance to the beginning of the assembly line from the end.
That means less resources wasted in production of faulty products that otherwise only get caught late in the production line
Taking it one step further, it also means that you can keep track of exactly which component items were mounted in which finished product items. So, in case of your sub supplier later identifying quality issues, you can do a "surgical recall", just recalling the exact individual products that contain faulty components, and no others.
Read more about authenticating components and the benefits it brings:
Reach Out To Us if You're Interested in Finding Out More About Industrial Product Authentication and How We Can Help You
Just send us a message here https://www.bluecromos.com/contact
Image by Smederevac on iStock