Bidirectional dynamic interaction model of seed traceability based on location analysis and its implementation

Abstract: Agricultural product safety concerns the national economy and the people's livelihood. From the starting point of agriculture production, the quality and safety of products, i.e. seeds, are especially important and need to be traced. To realize the traceability of crop seeds and help the production enterprise to analyze the operation status and realize the interactive marketing, in this paper, we constructed a bidirectional dynamic model of seed traceability based on location analysis. Through the seed electronic code, the model divided the traditional logistics information chain data into different sets of geographic codes, which were nested according to the geographical location of the distribution. The commodity seed from the factory to the user will experience a lot of path node, each path node dealer was required to writes its own geographic code. The geographic code contained the company information of the dealers at all levels and the place name information of the sale’s area. In the pre-sale stage of the seed, the geographic code set information was independent of each other. When the end user queried the seed electronic code, the flag was activated and the model started to dynamically analyze the location of the user. By converting the user's latitude and longitude information, this model received the detailed address information, started match the geographical code set of information from small to large, and ultimately found the dealer information or returned warning information. The query action of the user connected the geographical code sets of all levels, and the whole traceability data chain provided the traceability result to both the enterprise and the user at this time. The enterprise-side model consisted of a collection layer, a data sharing analysis layer, and an application presentation layer. When the enterprise obtained the traceability information, it can communicate with users, collecting user’s behavior data, and pushing enterprise marketing information. The enterprise-side model can be combined with supply chain data for data analysis to help improve and guide production. By selecting the data of sales and feedback of Huacheng 3366 wheat in ten counties, the goodness-of-fit was 0.997 8, which showed that the feedback from the code scanning could promote the sales. Among all the 14 092 barcode scanning position information, only 21 of them cannot be converted into the appropriate place name fields, and the model geo-code position conversion rate was around 99.851%, which can basically meet the business requirements of collecting and analyzing feedback information, helped enterprise effectively avoid the risk of transregional behavior. During the 36 months of trial operation, the monthly average number of server downtime was about 0.27 times, and the safe running days ratio was 99.087%, which can meet the requirements of enterprise traceability stability. The model was layered and independent of circulation information, which ensured the integrity, authenticity and security of traceability. The model added the bidirectional interactive data link on the traditional traceability model, so that the enterprise can interact with the user when scanning and obtain the user behavior data and push the marketing information. The model’s commissioning of the geocode location conversion rate, the number of days, the site safe operation and other technical indicators can effectively help companies master the actual operation, effectively prevent the occurrence of the transregional behavior, optimize the enterprise distribution channels, it also provides a more reliable means of information for the regulatory department to manage the seed market.