How counterfeiting can be solved with the AI monitored serialization technology
The solution to address the largest segment of merchandise counterfeited (ranging from $20 to $1000) need to be foolproof, and cost effective. The typical solutions offered today are holograms, rfid based tags, random 2D graph, scratchable QR codes, and Digital number based SMS system.
We found that every partial solution offered today to address counterfeiting can be counterfeited, meaning the tags or the scheme they employ can be copied and replicated easily. Below are our findings and rough estimate on how much it costs a counterfeiter to replicate and fool the solution provided.
Let’s explore each solution one by one:
All it takes to copy the hologram is, give one of the original to any hologram printer company and it will give you exact copies of it. It will even cost less to the Counterfeiters to get the exact Tags/Labels/Stickers from any hologram printing lab near them. Usually, the price ranges from less than 1¢ to 10¢ depending on the quantity, complexity, colors, layers, materials, and size involved.
Random 2D Graph stickers/labels
Some companies claim they generate or print certain shape or graphs on the label whose photo can be taken and sent to the cloud servers to authenticate. Unfortunately, they are still the same, no matter what the company claims about running a Machine Learning or AI to scan the image on the cloud. The truth is anyone who can take photos of the sticker and print them on a paper would pass the counterfeit-check. As long as the server gets the same scanned photo, regardless of the source (genuine sticker or copied sticker), it will pass the check.
RFID and NFC are no different than holograms when it comes to copying them. It is even easier, all you need is the RFID/NFC code set in one of the original item and set the same code in cloned RFID/NFC labels/tags, and you will be ready to pass the counterfeit check! Usual cost varies from few cents to a dollar based on type of RFID (active/passive) or NFC tags.
Unique Random Code per Product (QR or Barcode or number or text)
Companies, who provide the unique code per product, make more sense. There are two type of Codes usually provided:
A. Open (visible to everyone even before purchasing the product)
If the codes are open (not protected by some scratchable layer), then it is of not much use, anyone can copy them and pass the counterfeit check, no matter what company claims about the crypto tech, block-chain or advanced technology they use in the backend (although with the help of geo-location tech and AI, red flags can be raised for some cases).
Truth is, if you have the same code as the original, which you can see without buying the product, then you can copy it, put it on counterfeited product and pass the counterfeit-check.
Also, in the market, many say putting the supply-chain data on block-chain prevents counterfeiting – it is definitely not-true. Printing the code on the item means you are making the key to access the data on block-chain public, now any one can copy and replicate this code and pass the check. Block-chains are meant for totally different purpose and have different use-cases, and preventing the counterfeit of general goods is not really one of them.
Please note – by block-chain, we mean the block-chain in terms of technology, we have encountered many non-technology and business development people who confuse block-chain technology to something totally different depending on the context. To learn what really is a block-chain and how it can be made fast and cost effective, to be utilized for business-cases, you can check our whitepaper fast private blockchain.
B. Hidden or protected code
These are the physically protected QR or Barcode or Alphanumeric codes and you must buy the item before you can see the code, for example, scratchable card, and then verify with the backend server.
This does protect the item and if the backend technology is implemented properly, i.e. non-guessable codes are generated, then this will prevent the counterfeit.
But problem is, user has no idea before buying the item if it will be genuine or not, not only this company who has not implemented the backend properly may pass some items which may be fake, or may pass already scratched code again and again.
Secondly, if their DB is hacked then it will be very simple for the hackers to sell the codes to Counterfeiters.
The development and more importantly maintenance and security of such system is not simple and usually costs good maintenance cost. And the authenticating servers will have to be always online.
Introduction of SOLVE
Finally to provide better anti-counterfeit protection, we designed and implemented SOLVE.
We kept mainly 3 things in mind:
1. Foolproof but semi open (user must be able to check authenticity with decent probability before buying)
2. Secure (System must provide a backend security to the codes and the way they are accessed)
3. Simple to use for the end users and cost effective for the businesses
Foolproof and semi open
This is the most difficult part to design. After careful thoughts, we have introduced two tags which go together, one open (pre-purchase) tag and one protected (post-purchase) tag and each of them is unique for each product.
User can scan the pre-purchase tag before buying and get a good estimation of genuineness of the product and check any RED flags. And scanning the post-purchase tag (which is accessible only after purchase of the item) provides complete genuineness information.
There are two protection we have given to pre-purchase tag:
* This tag is tightly algorithmically coupled to the post-purchase tag. So if someone buys the product and scratches and scans the post-purchase tag, the pre-purchase tag’s status is also updated in our system. And any further scans of the pre-purchase tag fails the counterfeit-check.
If counterfeiter copies the pre-purchase tag and replicates it, the first scan of the post-purchase tag will invalidate all the products with corresponding pre-purchase tag
* Machine learning based anomaly detection system is always monitoring the scan activity of the tags. If the pre-purchase tag starts to randomly appear from the geographic locations where it should not be or if the patterns of scanning has any anomaly pattern our algorithms raise the RED flags. Which would mean all the products of the copied tag would be invalidated, a counterfeiter has to put new copied tags on products in retail stores, which would again be caught by our system within few scans.
Protection to post-purchase tag:
* Physical protection is provided to each tag, i.e they can only be scratched or opened once after purchase. Once they are scanned their status gets updated on our system and they can not be used again for authenticity check by any other device.
* They are algorithmically coupled with the pre-purchase tags. So any anomaly pattern on the pre-purchase tags will lead to RED flagging of the post-purchase tags too.
* Machine learning based anomaly detection system keeps monitoring the scan activity of the post-purchase tags and the pre-purchase tags. And immediately raises the RED flags for any abnormal or anomalous pattern.
Security of the auth system and tags are the most important criteria in designing the system.
Having the team with experienced ethical hacker, experience fighting hackers in gaming companies in past (Gala Net) and designing the secure systems for finance companies (Paypal) has helped not to overlook any potential cases where security can be a problem.
We can not expose more here, but our tags info is one way “scrypt” and private key secured so even if someone gets the DB of tags he or she can not get the tags code.
Simple to use for the end users and cost effective for the businesses
We have made sure the SOLVE can be scanned without need of any app download. This removes any friction from user’s side.
User can manage the purchases, warranty of the products and access to customer support with just a single scan.
Codes can be scanned by any scanner app, for iOS and new Android phones native camera works, also scanner of the popular apps, such as, WeChat, can be used.
The technology is cost effective, on the cloud and efficient to use so that it can be used by any company of any scale, without worrying about the cost. In most of the cases, the overall cost of our technology is usually lower than putting a hologram.
A single scan to check authenticity
A single scan to collect loyalty or participate in gaming
Overall Value Summary
A technology is useless if it can not be effectively put into use for mass adoption. While designing the SOLVE technology we have made sure it has short and long term business benefits.