In times of globalization, every day the volume of goods increases that is transported around the globe to enable international trade and serve customers with local and global products. Due to the durability of for example groceries or the urgency of spare parts, a timely delivery can pose a requirement for many supply chain networks. In current logistics, the limited transparency, siloed data storage and limited data sharing are widely spread issues in the shipment of goods along the supply chain.
The resulting blind spots and unavailability of data during the tracking of goods leads to a subsequent barrier to utilize statistical analysis or machine learning algorithms to perform precise delivery date forecasts, anomaly detection for fraud or damaged products, analytics based automated product quality management, or transport route optimization. The data storage on blockchains via the underlying distributed ledger technology, enables the tracking of goods without the limitation of a validating central and independent authority. It thus solves the issue of trust without relying on a central authority. In the remainder of the blog, we will list exemplary use cases arising from applying analytics to data stored in distributed ledgers highlighting a reference project.
Today’s logistic processes are complex and nested, thus various stakeholders are involved and suppliers and logistic providers along the supply chain will generally use their own IT infrastructure for tracking goods. As there is no neutral central authority to store data on logistic processes, every company will keep data inhouse and secure from other operators, which leads to blind spots during the tracking of shipments. It also prohibits any sort of data analytics on the entire supply chain, which, for example would be demanded for end-to-end supply chain optimization. The clear issue in this use case is trust and a distributed ledger can create trust between conflicting parties, while maintaining an independent record of shipments.
Figure 1: Data tracking in logistic processes for blockchain based data storage
For a product on its way to the store shelf, geolocations are recorded by all operating parties involved and all relevant tracking data, as displayed in figure 1, could be stored on the ledger. A shipment is equipped with a unique identifier and all participating parties that interact with the shipment, are required to accredit themselves on the ledger. This enables the distributed consensus to validate the location, time stamps and other information like for example status, temperature, or humidity and further offers the end user to track that operators have complied to all product regulations.
Also, the implementation of smart contracts (see Figure 2) between suppliers and merchants can solve another issue of trust. Currently, either the buyer must pay for goods up front or the seller must deliver products before payment, which requires trust between the two parties or an independent third party holding goods and money for a direct exchange. A blockchain can solve this issue by substituting the intermediate third party via a smart contract. The transaction is transformed through a peer to peer interaction and smart contracts regulate the record keeping of the exchanged goods and payments.
Figure 2: Smart contracts facilitate trust for value and good exchange between buyer and supplier
The unification of several sources of truth regarding data-points from the logistic processes provides a clear advantage for data analytics. First, the merkle tree generates a validity for the data stored and obligates all interacting parties to sign on the distributed ledger (Figure 3). Points of failure or fraud, can thus be traced back in detail and gaps in the supply chain can be quickly identified. This enables an early detection of high risk steps during the shipment process and suppliers and merchants can intervene. An increasing data quality, quantity and accessibility fosters better prediction accuracy in models forecasting the arrival of shipments and early warning algorithms for lack of compliance to product quality standards or fraud through fabrications. Based on that logic, advanced prediction or recommendation algorithms can be implemented – think of matching a personal diet to the nutrition values of food products. The transparent chain of information allows for including individual or ethical preferences like a large share of locally grown products. Once again, conducting analytics on a neutral, decentralized set of data allows for gain in information that lawmakers and regulators failed to establish to this point by introducing fuzzy and distributed food labels.
Figure 3: Ingredient information is added to a hash block on a Merkle tree to secure information
Currently solely few of these business models have been tested in proof-of-concept implementations and one can only imagine the disruptive potential and additive opportunities of logistic services relying on distributed ledger architecture for data storage at scale. Capgemini in collaboration with SAP is leading the way on increasing visibility in supply chain processes and aims at integrating the range of capabilities, including Internet of Things and Machine Learning, with blockchain technology in a single environment (read here for more information). Distributed ledger implementations in supply chain processes are excelling alternative centralized solutions due to the avoidance of conflicts of interests and will be particularly relevant for supply chain processes of valuable goods (to mitigate the risk of falsification and fraud), when regulations apply (e.g. trace bio products), or pharmaceutical products. Such or similar cases prove to be relevant to conduct further investments in supply-chain-management analytics.
This article was written in collaboration with my colleague Dr. Sebastian Olbrich.