This paper explores location-based product authentication in a situation where only the past locations of products that flow in a supply chain are known. We transform location-based authentication into a pattern recognition problem and investigate different solutions based on machine-learning techniques. The proposed solutions are studied with computer simulations that model the flow of genuine and counterfeit products in a generic pharmaceutical supply chain. The results suggest that machine-learning techniques could be used to automatically identify suspicious products from the incomplete location information.
The Auto-ID Labs have presented new research results at the first GS1 Global Standards Management Process - EPCglobal Joint Action Group meeting in Los Angeles.