Visibility data (providing details about supply chain activities in e.g. production, logistics, and quality processes) is of highly sensitive nature – not just in the food sector, but also beyond. Amongst other things, unauthorized data access can be (mis)used to uncover supply chain relationships, volumes, and other business context information. At the same time, it becomes increasingly ... [Show full abstract] important to share visibility data with trading partners, e.g. to meet customer requirements and legal obligations. So far, it is not a trivial matter to access or even discover that data, which is often stored in numerous distributed databases.
A possible means to overcome this predicament is a Discovery Service (DS), which has knowledge of the parties owning information about specific objects (e.g. product batches) and can provide pointers to the actual data sources to authorized clients while leaving no opportunity to misuse accessible data. It is important to note that a DS itself does not contain actual visibility data, but only references to it. Yet, even the knowledge that party A, B and C have information about a specific product is still sensitive as the querying client would be able to reveal the flow of goods and may take advantage of that knowledge. For instance, he could identify his supplier’s upstream vendor and, for the sake of saving costs, try to procure products directly from that upstream vendor rather than from his previous supplier. Hence, a DS should provide the ability to prevent such scenarios if it strives to achieve a high level of acceptance.
A key enabler for a DS is a trust-inspiring means to ascertain whether a querying party is actually part of a given chain of custody (CoC). In conjunction with a set of rules previously defined by the respective data owners, the DS is then able to decide which service addresses (if at all) can be provided in the message response to the query client. Afterwards, the querying party can gather detailed information about a specific object by querying the indicated data sources. In this context, our paper addresses the following research question: How can a DS ascertain if and to which extent a trading partner, whether known or not, is entitled to get pointers to visibility data stored in distributed repositories while ensuring privacy of the data owners?
In pursuing this research question, we first provide the relevant background including the current state of the art. Second, we explain and discuss the solution approach and complete our paper by outlining the solution we are going to realize in the course of the research project ‘FoodAuthent’ (funded by the German Ministry of Food and Agriculture).