O-KEY Revolutionising Product Authenticity with Nanotechnology and AI

Artboard 1
Artboard 1
Artboard 1
Artboard 1
Artboard 1
Artboard 1

Deliverables

  • UX and UI
  • Android Native Development
  • iOS Native Development
  • AI Training
  • QA
  • Project Management

In a groundbreaking approach to combat counterfeit products globally, Pufin-ID harnesses the power of Nanotechnology and Artificial Intelligence with their O—KEY solution.

With this solution one can label any physical product with a tamper-proof unique code, which cannot be replicated or copied. This unique code is called an O—KEY. The app is built to scan and verify the authenticity of any product with an O—KEY. The system consists of three main components:

  • An O—KEY, which consists of microscopic titanium particles in a random pattern embedded in either a QR code or directly on a product.
  • An App that can scan the O—KEY by utilising Pufin-ID’s Shortcut-developed SDK (Software Development Kit).
  • A server solution, managed by Pufin-ID, where all O—KEYs are registered, and where the physical product’s ID and associated metadata can be accessed via the SDK.

Shortcut’s task was to design and develop the app, including the iOS and Android SDKs, which can also be licensed by external Pufin-ID partners. We trained and deployed an AI model that enables the app to locate the O—KEY in a live-feed from the phone’s camera and to read the random particle pattern of O—KEY.

The magic of the system operates on multiple levels:

  • Every single O—KEY holds a random pattern, which cannot be forged or removed from the item.
  • The way the app is capable of identifying and decoding the O—KEY is unique.
  • The security and concept on the server are unique.

At Shortcut, we faced challenges in using the built-in capabilities of iOS and Android cameras, to create a user-friendly app that allows users to effortlessly to scan, identify, and decode an O—KEY. We tested various strategies, and the most powerful solution was to train an AI model to identify and analyse O—KEYs.

The AI model was trained using manual annotations on thousands of O—KEY images in order to make the app capable of recognising an O—KEY in various surroundings and lighting conditions. We chose the YOLO (You Only Look Once) framework. The model was then integrated into the O—KEY Verifier SDK and embedded into the app, enabling it to identify and isolate the O—KEY as required.

Want to
know more?

Contact me



Contact me

Mikkel Bay Ejlersen

Business Development Manager

Click to copy