Blockchain-enabled Data Sharing In Supply Chain

Information sharing among supply chain entities is vital for the success of a supply chain.For example,downstream retailers are usually in a better position to detect market trends,anticipate which products are going to break out,predict when product replenishment needs to be addressed in advance,and make response.Meanwhile,upstream suppliers and manufacturers have an information advantage in terms of product quality and quantity to supply.Such proprietary information,if properly shared,can reduce inventory level and lower operating costs for all supply chain members.Newly developed information technology,such as Internet of things,artificial intelligence,cloud computing,has promoted the development of supply chain information sharing.Despite extensive of the benefit of information sharing,the systematic implementation of information sharing across the board is still lacking in practice.Distrust,privacy concerns,data misuse,and the asymmetric valuation of shared data between entities often hinder data sharing.These problems call for a secure,efficient,fair,and trustworthy data-sharing mechanism.The key to such a successful system hinges on how to trace the data usage,determine the value of the seller’s data to the buyer and then compensate the seller accordingly.To address the challenges mentioned above,we design and implement a blockchain-enabled data-sharing marketplace for a stylized supply chain.It embraces the blockchain technologies that accommodate consensus,autonomy,and micropayments.We demonstrate how a blockchain can be used to overcome these impediments in supply-chain data sharing and provide a detailed tutorial with a stepby-step implementation for how to set up such a data exchange prototype using Hashgraph.We also explore the impact of quality information leakage and blockchain distributed ledger leakage in a supply chain system consisting of a manufacturer,a retailer,and a third-party certifier.This dissertation consists of seven chapters.Chapter 1 introduces the background,research significance,framework,and innovations.Chapter 2 reviews the related literature.In Chapter 3,We design and implement a blockchain-based data exchange to facilitate trustable information sharing in a supply chain.Our exchange design addresses several key challenges that often afflict supply chain data sharing,including privacy breach,data leakage,improper valuation of data,and unfair compensation.The exchange achieves transparency,security,fairness,and accountability through our proposed usage-based data valuation mechanism.We demonstrate how to deploy such a data exchange using Hashgraph and provide a detailed tutorial at the code level for the main functionalities in the Appendix.In Chapter 4,we propose a new usage-based data evaluation method,where the data to be shared are valued and priced based on their value expected to be generated according to the specific usage of the data.Within the context of valuing demand data in a newsvendor scenario,we derive expressions for the expected value of information(EVI)and the expected value of sample information(EVSI)when data can be viewed as independent draws from a distribution.The actual realized value may deviate from the initial set price.In Chapter 5,we extend Chapter 4 to evaluate multiple information provider’s contribution.We develop a principled framework to address data contribution in the context of supply chain usage model which is viewed as an algorithm.Given an algorithm trained on n data points from multiple data providers to produce an outcome,we propose DEA Data Shapley as a metric to quantify the contribution of each training data point to the outcome performance.To address computation complexity,we develop Monte Carlo algorithm to efficiently and approximately estimate DEA Data Shapley values in practical settings where algorithm is complex or number of data provider is huge.In addition to being fair and equitable,numerical experiments using a credit default dataset demonstrate that DEA Data Shapley has several other benefits: 1)the proposed metric can evaluate multiple target at the same time;2)it is more useful than the popular leave-one-out in providing insight on what data is more valuable for a given usage scenario;3)low DEA Data Shapley value data informs low quality data providers and outliers;4)high DEA Data Shapley value data suggest which data providers make major influence to improve the outcome.In Chapter 6,we investigate the impact of quality information leakage and blockchain distributed ledger leakage in a supply chain wherein the manufacturer can privately acquire the precise quality information of its product and freely choose whether to share it with downstream.We consider three different scenarios: information leakage(i.e.,there exists quality information leakage if the manufacturer chooses to acquire the precise quality information),partial protection(i.e.,supply chain use blockchain and encryption algorithm to ensure the safety of quality information,but the ledger is under exposure),and full protection(i.e.,both quality information and ledgers remain safe).We examine the effects of information leakage on the equilibrium strategies and payoffs of the manufacturer,the retailer,and the supply chain.It is shown that quality information leakage significantly reduces the manufacturer’s incentive to acquire and share the precise product quality information.Ledger exposure enhances supply chain transparency at the cost of manufacturer and supply chain’s payoff.Finally,we summarize our main insights,limitations and point out the future research opportunities in Chapter 7.We make several unique contributions to the data sharing literature in SCM.First,we make one of the first attempts to provide a blockchain-based solution to data sharing in SCM.We show how to leverage various blockchain features,such as decentralization,consensus mechanisms,cryptography,and micropayments,to address the challenges of data sharing.Decentralized structure precludes collusion or manipulation among sellers or buyers in advance;Data transactions powered by blockchain and cryptography provide a dependable ledger that is auditable for supply chain members.It motivates high quality information sharing in supply chain;Anonymity of the transaction protects user identity;Micropayment let dealing with transactions that are only worth pennies or less as.Second,we propose a novel usage-based valuation approach that determines the value of shared data by tracking how the data are used in the buyer’s model and how much improvement it contributes to model performance.We develop the EVSI and EVI methods to accommodate different data-sharing scenarios.These methods can utilize data and model at the same time and generate a fair value estimation which will enhance member consensus.For value attribution and distribution,the data exchange tracks and verifies the revenue or loss generated from using the data and then compensates the buyer and seller according to the terms prespecified by the two parties.The final settlement is only completed after the actual use of the shared data is accomplished and the usage value is realized and observed.New methods provide a win-win solution for supply chain information sharing.Third,for the first time we investigate the impact of quality information leakage and ledger information disclosure in a supply chain system.Quality information leakage analysis informs the negative impact of information leakage to supply chain member’s decision.Ledger information leakage analysis suggests the side effect of using blockchain in supply chain.Both analyses provide guidance for supply chain information sharing.Fourth,we provide a closed-loop solution on how to implement the proposed mechanism using Hashgraph.We demonstrate how to deploy such a data exchange using and provide a detailed tutorial at the code level for the main functionalities.This dissertation provides both theoretical and technical solution which address vital problems in supply chain information sharing.It will boost supply chain efficiency and promote the application of blockchain and other IT technology in supply chain management.This dissertation has a strong forward-looking and positive social value.