Agro-food collaborative traceability platform for distributed environment

Abstract: With the development of information technologies and the improved consciousness of food safety, various agricultural product traceability systems have been developed and applied in China. Lack of data share between different systems has led to the decreasing of the credibility and analysis capacity. Therefore, establishing a collaborative traceability platform on the distributed environment is an urgent need to enhance the ability of unified traceability and collaborative analysis. Combining data from different business systems, a platform framework was constructed, which included 3 layers i.e. data integration, model analysis and application supply. Enterprise credit evaluation model was the important part in the model analysis layer. An evaluation index system with 4 factors and 14 sub-factors was designed. The index included enterprise qualification, quality inspection, product traceability, and consumer feedback. The evaluation factors' weight was confirmed using the AHP (analytic hierarchy process) method and the evaluation score was calculated with the weighted sum algorithm.. Multi-source heterogeneous data integration technology was researched based on ontology. It included 3 parts: Data format integration, semantic mapping and heterogeneous data analysis. Traceability information was described with five-factor knowledge ontology and expressed with OWL (Ontology Web Language). Layered dynamic tracing technology was implemented in order to trace different agro-food in a unified platform. In this part, there were 2 layers including basis information and extended information designed, and the dynamic mechanism of invoking the traceability data was developed according to the product features. Taking Tianjin City as an example, and using Microsoft Visual Studio 2010 as the system design and development tool, the united traceability platform of agro-food safety was developed. It provided traceability service for the consumer, which was production tracing, enterprise credit overview and production evaluation. The platform also provided data analysis function for the government, which was base registration statistics, farming process statistics, traceability barcodes print amount analysis, agro-food price analysis and enterprise credit evaluation. The ability of concurrent access and the retrospective effect were analyzed. The results showed that when the number of concurrent users reached 40, the average response time of a single transaction reached 19 s. The comparison of the platform and the other 4 traceability systems was performed. The results showed that this platform presented the features of multi-category tracing, tracing information diversity, and interactive features. In the future work, the platform performance will be optimized through the updated hardware and software. The decision support function will also be enhanced through developing the model on the accumulated data. The research achievements provide the beneficial support to improve the integration level of traceability information and analysis capacity of traceability results.