Counterfeit products, particularly pharmaceuticals, electronic devices, and apparels, are widespread. They threaten consumer safety and cause huge economic losses to licit supply chain partners and governments.
Although a number of traditional anti-counterfeiting technologies, such as holograms and chemical tags, are available to combat counterfeiting, they are vulnerable to imitation or being reused. Besides, these technologies are intended to protect individual items, rather than to safeguard an entire supply chain. As such, fake products may likely be injected into the supply chain to hurt end-consumers.
Track-and-trace technology based on Radio Frequency Identification (RFID) has recently emerged as a promising tool to combat counterfeiting, because of its automatic and non-line-of-sight capability to identify massive product items. By maintaining an electronic pedigree (e-pedigree) that records the transaction information of product items along the supply chain, this approach stands out for protecting the supply chain against infiltration, eliminating theft and fraud, facilitating recall of defective products, and supporting remote authentication.
However, a number of technical and critical issues have yet to be solved for practical implementation of RFID-based track-and-trace anti-counterfeiting. These include generation of accurate initial product e-pedigree in fast moving manufacturing lines, precise e-pedigree updating in batch product distributing and receiving, and fast e- pedigree queries for remote and real-time product authentication from end-customers. Without fully addressing these issues, the accumulated product e-pedigree data would be untrustworthy, rendering any subsequent operations of track-and-trace and product authentication unreliable.
This thesis investigates the crucial implementation issues in RFID-based track-and-trace anti-counterfeiting. It firstly presents an innovative track-and-trace anti-counterfeiting system, based on which a TDPS algorithm is proposed for generation of initial product e-pedigree in fast moving production lines. The TDPS overcomes many practical issues, such as tag writing error and tag locking failure, and helps identify the bottleneck of initial product e-pedigree generation. To tackle the bottleneck, the TDPS is further optimized by incorporating a block writing method to enhance the tag EPC writing efficiency and an integration method to balance the overhead of RFID equipment.
In product distributing and receiving, a mechanized 3D scanning method is proposed to improve bulk item identification rate and enhance the accuracy and completeness of product e-pedigree. Indeed, RFID-based track-and-trace anti-counterfeiting mandates a relatively high bulk item identification rate for product authentication and e-pedigree updating. Experimental results demonstrate that the mechanized 3D scanning can achieve a bulk item reading rate of up to 98.9%, which largely outperforms the widely documented bulk reading rate (70%) in real applications.
In retailing level, the efficiency of e-pedigree queries would hugely impact on customer shopping experience and the effectiveness of track-and-trace anti-counterfeiting. A partition-based method is therefore developed to cluster product e-pedigree data to improve the speed of e-pedigree queries. This approach partitions the accumulated e-pedigree data into fixed and dynamic groups, such that queries are conducted mainly on active data, rather than on the whole historical data sets.
By addressing the above key issues, this thesis contributes to making implementation of RFID-based track-and-trace anti-counterfeiting practically viable and reliable. / published_or_final_version / Industrial and Manufacturing Systems Engineering / Master / Master of Philosophy
|Creators||Yang, Bo, 楊波|
|Publisher||The University of Hong Kong (Pokfulam, Hong Kong)|
|Source Sets||Hong Kong University Theses|
|Rights||The author retains all proprietary rights, (such as patent rights) and the right to use in future works., Creative Commons: Attribution 3.0 Hong Kong License|
|Relation||HKU Theses Online (HKUTO)|
Page generated in 0.0027 seconds