Research on a Robust Traceability Method for the Assembly Manufacturing Supply Chain Based on Blockchain

Abstract

The management of assembly manufacturing supply chains in a cloud computing environment poses various challenges, including extensive regional management, a lack of transparency in the supply chain, an absence of a secure and effective traceability mechanism, and difficulties in achieving safe traceability. Therefore, this paper proposes a robust traceability scheme for assembly manufacturing supply chains based on blockchain technology. The solution utilizes IoT devices to collect data on product production and processing while ensuring the security and privacy of traceability information through digital signatures and hash encryption algorithms. Furthermore, by employing an “on-chain + off-chain” mixed storage strategy, the scheme achieves secure storage of traceable data. Additionally, the proposed scheme enhances the reliability of the traceability process through an efficient on-chain query mechanism and an off-chain trusted verification method. This research provides both theoretical foundations and technical pathways for enhancing the reliability of assembly manufacturing supply chains as well as their practical application.

1. Introduction

Cloud manufacturing represents a networked manufacturing model that seamlessly integrates cutting-edge technologies including cloud computing, the Internet of Things (IoT), and Artificial Intelligence (AI). It effectively transforms conventional manufacturing resources into versatile cloud-based services accessible via a dedicated platform [1]. The advent of this paradigm has significantly contributed to enabling multi-agent business collaboration mechanisms. By leveraging the power of the aforementioned technologies within a unified framework offered by the cloud platform itself, users are empowered to establish comprehensive supply chain networks spanning multiple enterprises and geographical regions—thereby fostering streamlined cooperation across diverse domains encompassing design innovation processes alongside production management logistics, among others. Nevertheless, operating within this dynamic ecosystem entails managing vast amounts of resources, forming an information pool involving numerous participants and complex product life cycles.

The objective of supply chain traceability [2] is to establish a comprehensive management system that encompasses the entire process from production, processing, and transportation to sales, enabling effective information monitoring from origin to market end. In the cloud computing environment, the collaborative assembly manufacturing supply chain with multiple agents can achieve centralized and transparent management of each participant, ensuring real-time monitoring and traceability at every stage. However, this model also faces several challenges and areas for improvement: (1) The extensive scope of supply chain management involves numerous upstream and downstream enterprises, resulting in difficulties for different parties in constructing product information and exchanging/integrating data, leading to the generation of information islands [3]. (2) Due to the centralized supply chain management model, most enterprise production information is stored in a vulnerable cloud environment which exposes it to security threats like hacker attacks, thereby jeopardizing data integrity and availability [4]. (3) There is no guarantee that shared user information aligns with actual situations; false sharing significantly damages both reputation and efficiency within the cloud manufacturing service system.

Blockchain technology relies on decentralized bookkeeping, distributed storage, consensus mechanisms, and cryptographic principles [5]. Its unique chain data structure ensures the inclusion of the previous block’s hash value in each block, thereby providing a highly traceable and data-integrity-driven system [6]. This characteristic offers innovative ideas and solutions for addressing various challenges encountered in traditional traceability methods.

In recent years, scholars both at home and abroad have been actively engaged in research on the integration of blockchain technology with diverse application scenarios to address practical needs in the real world. In terms of data privacy protection and secure storage, Gao et al. [7] addressed the issue of medical data leakage by using hash algorithms and symmetric encryption techniques to encrypt patient identity information and medical records, respectively, and introduced a ring signature mechanism to achieve anonymity and traceability of hospital signatures. Sun et al. [8] proposed a multi-chain storage architecture based on blockchain smart contracts to address the problems of large cross-regional span, long cycle, and massive heterogeneous data in the traceability process of fruit and vegetable products, thereby achieving trusted evidence storage of traceability data from planting and processing to the storage of agricultural products. Lai et al. [9] proposed a secure medical data sharing scheme based on blockchain to meet the data sharing needs among medical institutions. They encrypted privacy data using a traceable ring signature algorithm and combined it with an access control mechanism and Sequential Control Organization (SCO) to ensure the secure circulation of cross-institutional data. Singh et al. [10] designed an improved trust and accountability mechanism data sharing framework to address the challenges of cross-domain secure communication in the industrial Internet of Things. They utilized a cloud platform to connect distributed gateways and data receivers and ensured reliable data sharing through user identity authentication. In the field of data traceability, Xiong et al. [11] applied blockchain technology to the information traceability scenario of vaccines, achieving the information on-chain and traceable supervision of the entire process from circulation to vaccination based on the FISCO BCOS platform (v2.9.1). Li et al. [12] proposed a master–slave multi-chain structure information management model for the traceability of the grain supply chain based on the data characteristics of each link in the supply chain. They used the Raft consensus mechanism and an improved Proof of Work (PoW) algorithm to reach consensus on the chain and achieved the automatic on-chain and efficient traceability of information through smart contracts. In terms of optimizing the storage and query performance of blockchain, Ren et al. [13] designed a secure storage mechanism for spatio-temporal data called BSMD, which effectively alleviated the storage pressure of the blockchain system by integrating on-chain anchoring and off-chain storage resources. Ye and Park [14] proposed a vehicle data secure storage system combining blockchain and IPFS, allowing users to access and manage vehicle-related data through a decentralized application (DApp) on the Ethereum platform (Berlin hard fork).

The above-mentioned research indicates that blockchain technology has been widely applied in multiple fields such as the Internet of Things (IoT) communication, healthcare, and food safety. Whether it is reliable data exchange among IoT devices, privacy protection of medical records, or traceability and supervision throughout the food production process, blockchain technology has demonstrated significant technical advantages and broad application prospects. Moreover, the distributed resource sharing mechanism under the cloud computing model shares certain commonalities in architectural characteristics with blockchain. Integrating blockchain technology deeply into the actual application scenarios of the manufacturing sector not only aligns with the development trend of digital transformation in manufacturing but also helps enhance system credibility and collaborative efficiency. Therefore, an increasing number of scholars are focusing their research on the application of blockchain in manufacturing.

In the field of supply chain traceability, Ho et al. [15] addressed the complex and inefficient traceability process of aircraft parts supply chains by building an efficient traceability system for them based on the Hyperledger Fabric platform (v1.2), achieving full traceability of the flow information of aviation parts among manufacturers, maintenance plants, and airlines. Zhu Ziwei et al. [16] combined blockchain technology with RFID, Internet of Things (IoT), and QR code technologies, setting up three independent ledgers to record vehicle identification information, parts procurement and configuration information, and vehicle logistics information, thereby realizing the transparency and traceability of the automotive manufacturing supply chain. Kuhn et al. [17], in light of the characteristics of the automotive manufacturing network, proposed a blockchain-based traceability framework for autonomous vehicle electronic systems, aiming to meet the demand for transparent data sharing during the manufacturing process of electronic systems. Cohen and Rozenes [18] designed a digital traceability system for complex assembly products that integrates blockchain and smart contracts and introduced digital twin technology to achieve full traceability of the bill of materials, assembly activities, and related resources of multi-level assembly products in intelligent workshops. In terms of data management, Zhu Jian et al. [19] proposed a blockchain-based data traceability solution for the industrial Internet, using RFID technology to collect the full life cycle data of industrial products and storing key data on the blockchain through symmetric encryption algorithms to ensure the secure storage and trusted traceability of industrial data. Hasegawa and Yamamoto [20] constructed a blockchain-based IoT data management platform, using smart contracts to uniformly manage data from IoT devices such as smartphones and sensor nodes and ensuring the consistency and reliability of data through the integrity verification mechanism of the blockchain. Through a systematic analysis of the above studies, it can be seen that in recent years, the academic community has mainly focused on the integration of traditional traceability technologies and blockchain technology, actively exploring their application value in various supply chain traceability scenarios. At the same time, there have been initial explorations into the traceability issues of assembly manufacturing supply chains in cloud environments, with some research results concentrating on the optimization and improvement of blockchain in terms of secure storage and trusted traceability. However, there are still several urgent problems to be solved in the research of assembly manufacturing supply chain traceability in cloud environments. Based on the review of the current research status, this paper will systematically elaborate on the existing deficiencies.

(1)

The assembly and manufacturing supply chain involves multiple production stages. Under the cloud model, manufacturing enterprises are widely and highly dispersed geographically. Each enterprise usually stores production information independently in local databases, lacking unified data standards and cross-enterprise information sharing mechanisms. This leads to a severe “information island” phenomenon, which restricts the effective improvement of the overall traceability capability of the supply chain.

(2)

In the cloud mode, manufacturing enterprises generally rely on data collection devices to integrate multi-source heterogeneous production data and manage it centrally through a centralized database. They also achieve information interconnection and interoperability through cloud platforms. However, such a centralized architecture is vulnerable to malicious attacks during data storage and transmission, posing security risks such as data leakage, tampering, or loss, and thus is difficult to meet the high-trustworthiness requirements for data management.

(3)

In the distributed collaborative manufacturing model, as the manufacturing process of products continues to advance, each manufacturing node continuously generates a large amount of streaming data. There is an urgent need for strong data storage capabilities and efficient retrieval mechanisms to extract key information. Currently, the on-chain query function of blockchain systems is relatively limited, and the retrieval efficiency is low. Especially when quickly locating specific production stages or equipment status, the response is slow, which may lead to delays in obtaining key information and thereby affect the real-time monitoring and decision-making efficiency of the manufacturing process.

In summary, the multi-agent collaborative assembly manufacturing supply chain based on cloud platforms faces problems such as the risk of data loss due to centralized storage, the possibility of tampering with stored information, and insufficient credibility of traceability results during the traceability process. Blockchain technology, with its distributed storage, efficient consensus mechanism, and privacy encryption features, provides a new technical path for the storage of production information, secure order transactions, and reliable data traceability in the assembly manufacturing supply chain in a cloud environment. Therefore, this paper, based on the principles and architecture of blockchain technology, conducts research on the secure and reliable traceability of the assembly manufacturing supply chain in a cloud model, which has significant theoretical value and practical significance. The main research contents of this paper are as follows:

(1)

A research framework for a trusted traceability method of assembly manufacturing supply chain based on blockchain is constructed to address the problems of centralized storage being prone to tampering and low efficiency in cross-enterprise data traceability in the traceability process of the assembly manufacturing supply chain under the cloud mode. Starting from three dimensions: secure storage of traceability data, consensus mechanism of the traceability network, and efficient traceability verification, a trusted, efficient, and auditable traceability of the assembly manufacturing supply chain in the cloud environment is achieved.

(2)

In light of the operational characteristics of the assembly manufacturing supply chain in the cloud environment and the demand for secure data storage, this paper designs a traceability system for the assembly manufacturing supply chain based on a master–slave multi-chain architecture, and uses the Channel isolation technology to protect the data privacy among different business links. On this basis, a blockchain-based traceability data security storage model is proposed: the SHA-256 hash algorithm is adopted on the chain to encrypt the sensitive information of enterprises, and the ECDSA digital signature algorithm is combined to authenticate the on-chain data, ensuring the authenticity and non-repudiation of the data source; while off-chain, detailed traceability content is stored through traditional databases, achieving the collaborative storage of on-chain metadata and off-chain detailed data. This solution effectively integrates the tamper-proof advantage of blockchain with the high-capacity storage capability of traditional databases, ensuring the security and integrity of core data while alleviating the storage pressure of the blockchain system.

(3)

In response to the practical demand for efficient traceability in industrial scenarios, this paper proposes a high-efficiency traceability verification method for assembly manufacturing supply chains based on a blockchain and a hierarchical storage architecture. In the main traceability chain, the Bloom filter in the B-Merkle structure is utilized to screen out the problem links related to the target traceability code and extract the corresponding block index. In the sub-chains, a Skip List multi-level index structure is introduced to accelerate the acquisition of specific traceability content through a parallel retrieval mechanism, significantly reducing query latency. Off-chain, the elliptic curve digital signature algorithm is used to verify the hash value of the on-chain data and compare it with the hash value of the original off-chain data to complete the dual verification of the integrity and authenticity of the traceability data. Experimental results show that the proposed traceability verification method has high retrieval efficiency and good scalability, and can effectively meet the actual demand for rapid and accurate traceability in assembly manufacturing supply chains.

2. Related Technologies

2.1. Blockchain

The blockchain technology is a novel application mode that integrates distributed data storage, point-to-point transmission, multi-party consensus, cryptography, and other computer technologies. It possesses the characteristics of decentralization, security, transparency, and immutability [21]. Leveraging its unique chain structure and based on chronological order, hash pointers are connected to form the longest main chain which ensures traceability and positioning of data within the chain. The block structure is illustrated in Figure 1. According to different levels of openness and application scenarios, blockchain can be categorized into three types: public chain, private chain, and consortium chain [22]. Among them, the public chain exhibits the highest level of decentralization as any user can participate without permission, and all data is openly accessible to the entire network. The public chain also ensures a high degree of decentralization where no user authorization is required; nodes freely join or exit with anonymous identities while maintaining complete transparency of on-chain data. The consortium chain reduces the degree of decentralization to some extent by allowing only authenticated nodes to access the network, requiring node identity verification. On the other hand, the private chain is primarily deployed within an organization where management and control rights are typically held by that organization itself. It serves mainly for internal application scenarios.

2.2. Consensus Mechanism

The consensus mechanism is an algorithm that guides nodes in a blockchain network to maintain consistency, ensuring the verification of ownership of accounting nodes and accuracy of data [23]. By establishing a set of conditions and rules, the consensus mechanism accurately identifies trustworthy behavior nodes and ensures data consistency and synchronization through processes such as consensus decision-making, block creation, transaction confirmation, and sequencing. Implementing the consensus mechanism not only enhances the security and legitimacy of transaction information but also strengthens the overall robustness of the system. Commonly used consensus mechanisms include Proof-of-Work (PoW), Proof-of-Stake (PoS), Delegated Proof-of-Stake (DPoS), and Practical Byzantine Fault Tolerance (PBFT). The corresponding algorithms are presented in

Table 1. Comparison of consensus algorithm performance.

	PoW	PoS	DPoS	PBFT
Energy Consumption	High	Moderate	Low	Low
Latency	High	High	Low	Low
Fault Tolerance	50%	50%	50%	33%
Throughput	Low	Low	High	High
Scalability	Good	Good	Fair	Poor
Applicable Environment	Public Chain	Public Chain	Consortium Chain	Consortium Chain

.

2.3. Improved EG-PBFT Consensus Algorithm

The assembly manufacturing supply chain involves multiple participants such as suppliers, manufacturers, and logistics providers. In response to the data sharing requirements in this scenario, this paper proposes an improved group election-based EG-PBFT consensus algorithm. The traditional PBFT algorithm requires two rounds of full network broadcasts to achieve the consistency confirmation of the operation sequence by replica nodes. However, in the process of traceability data storage in the assembly manufacturing supply chain, it is only necessary to ensure the consistency and validity verification of data among nodes to achieve the consensus goal. Additionally, in the first round of the broadcast phase of the PBFT algorithm, all secondary nodes have already received the request message released by the primary node. Based on the actual requirements and communication characteristics of the above application scenarios, this paper optimizes the traditional consensus protocol in combination with the requirements for efficient and scalable consensus mechanisms in the assembly manufacturing supply chain. The execution process of the improved EG-PBFT consensus algorithm is shown in Figure 2.

2.4. Smart Contract

A smart contract is a decentralized, self-validating computer transaction protocol that consists of program code and stored files. These programs are automatically executed by specific nodes based on predefined conditions [24]. As the middleware interacting with the underlying blockchain and application system, it is characterized by data transparency, immutability, and long-term operation [25]. Since smart contracts do not require involvement from third-party authorities or centralized agent services for execution, they can promptly respond to user requests. Once deployed, the contents of the contract remain unaltered.

3. Design of a Traceability Framework for Assembly Manufacturing Supply Chain

With the aid of blockchain technology, this paper employs a four-layer architectural design to establish the comprehensive framework for traceability in assembly manufacturing supply chains, as illustrated in Figure 3. The architecture primarily comprises a hardware support layer (data acquisition layer), data storage layer, business interaction layer, and user layer. Users engage with the data via the blockchain network utilizing smart contracts as an intermediary for interaction, aiming to ensure openness, transparency, security, and trustworthiness throughout the intricate traceability process of assembly supply chains.

3.1. Hardware Support Layer (Data Acquisition Layer)

The hardware service layer, serving as the cornerstone of the system’s data source, is responsible for collecting and inputting product traceability information. By utilizing cloud servers, IoT control terminals, sensors, and other devices, discrete manufacturing nodes are sensed and collected to be uploaded onto the blockchain system in order to generate the necessary traceability data required by the traceability system.

3.2. Data Storage Layer

The data storage layer constitutes the fundamental component of the traceability data storage system for assembly manufacturing supply chains, encompassing both off-chain databases and the Fabric network. Upon entering the storage layer, data undergoes encryption via a hash algorithm and is subsequently marked with a digital signature. The encrypted data is then packaged and uploaded to the blockchain using an enhanced packet consensus algorithm, ultimately storing complete traceable information within the off-chain database.

3.3. Business Functional Layer

The business function layer primarily offers functionalities such as user authentication, rights management, information entry, data synchronization, traceability query, and others. Users select the corresponding service based on their specific requirements while the backend executes transaction requests upon receiving user operations and returns a data visualization interface to facilitate front-end and back-end interaction. The backend accomplishes traceability tasks by invoking smart contracts.

3.4. Application Layer

As a window for the system to provide users with interaction, the user layer is responsible for receiving the data input by users, and passing it to the business interaction layer or data storage layer for processing and output, which provides the data basis for system operation and calculation.

4. Key Technologies for Achieving Traceability in Assembly Manufacturing Supply Chains

4.1. The Architecture of Multi-Chain Traceability with a Master–Slave Structure

The assembly production process involves intricate business operations, multiple participating enterprises, a vast array of accessories, and a mixed flow production system, which ultimately leads to low efficiency in traceability within the supply chain. Blockchain technology enables secure information storage in a decentralized network and establishes an immutable and traceable historical transaction record [18]. In light of this, this paper introduces blockchain technology as a secure and trustworthy platform for information sharing among all participants across the supply chain. Depending on the type of interconnection required, blockchains can be categorized into single chains, side chains, and interconnected chains. The traditional single-chain model proves to be inefficient in handling multiple digital assets and fails to meet the traceability requirements of complex supply chain environments. In contrast, the master–slave multi-chain structure effectively safeguards enterprise private data by utilizing digital certificates for node access control and implementing data isolation technology. Therefore, this paper proposes the development of a master–slave multi-chain structure-based traceability system for the supply chain in the assembly manufacturing industry. The specific architecture is illustrated in Figure 4.

The paper proposes a methodology to decompose the intricate assembly structure into multiple core sub-assemblies and sequentially record the data of each assembly node based on the assembly sequence. The process encompasses the final assembly node, logistics node, and sales node, which collectively constitute the core traceability chain. To ensure data privacy protection across all supply chain links, this paper introduces access control technology and establishes dedicated slave chains for each stage of assembly, final assembly, logistics, and sales. Each slave chain is responsible for storing key traceability data specific to its corresponding stage. Through seamless collaboration between the main chain and slave chains, the system accurately locates data and achieves end-to-end traceability from raw materials to finished products. This guarantees secure data storage and fosters complete trust in traceability. In consideration of the sensitive multi-party data involved in the traceability process of assembly manufacturing supply chains, and the specific access requirements of different participants, this paper selects an alliance blockchain as the preferred type to achieve the aforementioned functionalities. With its distinctive authorization mechanism, the alliance blockchain can effectively restrict data visibility and access, provide a higher level of security for supply chains, facilitate accurate sharing and management of data, thereby assisting enterprises in achieving secure storage and trusted traceability of assembly manufacturing supply chain data.

4.2. Hierarchical Storage Model

The blockchain’s distributed storage mechanism ensures that any written data is replicated across all nodes, providing a guarantee of immutability and reliability. However, the traceability process in assembly and manufacturing supply chains generates massive amounts of data. Uploading this data directly to the blockchain would not only reduce network efficiency but also significantly increase system maintenance costs. The paper proposes a layered storage architecture that combines off-chain database and blockchain to address the performance bottleneck caused by high storage demand, thereby achieving secure and traceable data storage. Figure 5 illustrates the data storage model of assembly manufacturing supply chain traceability based on this architecture.

According to the object-oriented nature and the degree of importance in traceability data, it can be categorized into two types: publicly disclosed key traceability data and private data exclusively for enterprises. Key traceability data refers to information that consumers are most concerned about when purchasing products, such as basic product details and quality inspection results. On the other hand, private data includes sensitive information like technical specifications and process designs. In case of product quality issues, auditors may access private information to analyze specific causes of impact. The division of traceability data for each stage is presented in

Table 2. Table of traceability data division for each link.

Supply Chain Stage	Key Traceability Data	Privacy Data
Component Tracing	Component supplier, component product information, and quality inspection status	Component processing information
Assembly Tracing	Assembly manufacturer, assembly parameters, and quality inspection status	Assembly production information
Logistics Tracing	Transport company, transport route, and arrival quality inspection status	Other storage and transportation information
Sales and After-sales Service	Sales company, transaction number, and repair order number	Other sales and after-sales service information

.

In order to ensure the traceability of product quality, enterprises in the supply chain must organize key data that impacts quality during the production process into transactions and submit them to the corresponding enterprise slave chain. Once the data is packaged through consensus across the entire network, the unique identification of each block is synchronized with the main chain network via a cross-chain channel. The specific traceability data from the chain is stored, and the main chain assumes responsibility for maintaining all information from the chain. During the traceability process, the system can quickly retrieve stored content through main chain positioning and chain traceability. Figure 6 illustrates a master–slave multi-chain block structure.

The users register with the Certificate Authority (CA) to acquire digital certificates and establish a connection between their legal identity and rights. In the “blockchain + database” hierarchical storage mode, the original traceability data is stored in an off-chain database, while enterprise privacy data undergoes SHA256 algorithm-based hashed encryption before being added to the chain. Additionally, on-chain data is signed using the ECDSA algorithm to ensure its searchability and enable mutual verification between on-chain and off-chain sources. This approach leverages blockchain technology to securely store data while alleviating pressure on blockchain storage capacity. Simultaneously, the underlying off-chain database offers ample capacity, rapid response times, and robust scalability to cater effectively to customized business requirements.

4.3. Efficient Traceability Verification Model

In light of the common challenges encountered in the cloud-based assembly and manufacturing supply chain, such as the absence of a secure and reliable traceability mechanism, limited credibility of traceability information, vulnerability to single point failures due to centralized storage systems, and inadequate efficiency in information transmission, this paper proposes an integrated approach for verifying traceability in assembly and manufacturing supply chains that prioritizes both security and efficiency. This approach is based on a hierarchical storage architecture and incorporates advanced technologies including blockchain technology, hash algorithms, bloom filters, and digital signatures to enable efficient retrieval of on-chain data while ensuring secure verification of off-chain data. The specific structure of the proposed model for efficient traceability verification in assembly manufacturing supply chains is illustrated in Figure 7.

4.3.1. Efficient Retrieval on the Blockchain

The introduction of built-in index technology as the core of the efficient query algorithm aims to enhance traceability query efficiency in the assembly manufacturing supply chain. By utilizing the B-Merkle structure, the main chain swiftly selects location blocks containing traceability sources, while parallel querying of the target block’s traceability information set is achieved through jump table indexing from the chain. This enables comprehensive traceability across all links within the assembly manufacturing supply chain.

(1)

Block structure based on B-Merkle tree

The blockchain is composed of a series of cryptographically linked data blocks, with transactions stored in a hierarchical structure known as the Merkle tree. The tracing query begins from the current block and traverses down the Merkle tree until it finds the target data, continuing to traverse previous blocks if necessary. However, this querying process has low efficiency. On the other hand, Bloom filter is a spatially efficient probabilistic structure [19] that enables quick determination of whether a target element exists in a search sequence.

When a new element is added to the bitmap, all positions are set to 0. Then, the hash function is applied through k unbiased functions and k corresponding bits are set to 1 on the bit array based on the hash value for indexing purposes. During searching, the output k value is evaluated. If all k values are 1, it suggests that the query object may exist; if any of them are 0, then it confirms that the element does not belong in the set. Therefore, integrating Bloom filter into Merkle-Tree enhances on-chain retrieval efficiency and reduces computing resources consumption. The B-Merkle structure is illustrated in Figure 8.

(2)

Fast query based on Skip List

The blockchain is a singly linked list sorted chronologically and composed of hash pointers in series. When performing on-chain queries, it typically starts from the latest block and traverses backwards based on the hash value until the desired result is found. However, this approach results in slow query efficiency. On the other hand, Skip List is a randomized data structure that enables fast insertion, deletion, and querying through parallel linked lists with improved time complexity. Therefore, to achieve efficient on-chain retrieval, the concept of jump table has been introduced into blockchain by employing a multi-stage index mechanism using skip tables. Figure 9 illustrates the traceability blockchain jump table retrieval structure.

(3)

Efficient on-chain retrieval

The traceability request is initiated by the user, and the traceability query contract is invoked by the background. During the querying process, the system sequentially examines each block’s Merkle root on the main chain based on the submitted traced source code from the user. By utilizing feedback from a Bloom filter, it identifies the block containing the traced source code and extracts the slave chain block hash stored within that block. Subsequently, using this chain block hash and jump table index, it swiftly locates and retrieves specific trace content from the target block. Figure 10 illustrates this efficient retrieval process on an assembly manufacturing traceability chain.

4.3.2. Off-Chain Data Verification

The chain incorporates a stringent data verification mechanism to ensure the integrity and authenticity of locally stored data. By leveraging the complete traceability information associated with the on-chain transaction hash, off-chain data is authenticated through hash comparison, while ECDSA signature verification determines the legitimacy of the data source and detects any potential tampering during transmission.

(1)

The inspector enters the unique traceability source code ZID of the product, retrieves the traceable transaction set T_c from the blockchain, and extracts the transaction hash address, digital signature $Sig\_I$ = ECDSA(h(M),d_M), and the enterprise’s public key Q_m.

(2)

The plaintext information corresponding to the T_c in the local database should be queried based on the transaction hash address stored on the chain. If the transaction address exists, detailed traceability information N can be extracted from the local database.

(3)

The signature (r,s) is verified to be an integer within the interval [1, n − 1]. If it meets this criterion, Hash(N) is computed using the same hash algorithm employed in the signature step for local detailed traceability information N.

(4)

The inspector verifies the Q_m digital signature $Sig\_I$ by validating it against the public key provided by the enterprise, in order to ascertain the reliability of the message source. Initially, input message summary h(m), signature value public key Q_A(x,y), signature(r,s) and base point G(x,y); Calculate w = s⁻¹modn, u₁ = (h(m) × w)modn, u₂ = rwmodn, and R = (xR,yR) = u₁G + u₂Q_A. If R = 0, then the signature is deemed invalid and rejected. If v = xRmodn, only when v = r holds true can it be proven that the corresponding private key generated the signature and thus affirming the trustworthiness of the data source. Upon successful verification, Hash(M) is returned.

(5)

The consistency between Hash(M) and Hash(N) determines the integrity of data N in the chain; if they match, it verifies that N has not been tampered with, whereas any discrepancy indicates a change in N and renders the traceable data unreliable.

5. Traceability Scheme Implementation

The reliable traceability scheme of assembly manufacturing supply chain based on blockchain is divided into the processes of information collection, encryption storage, and traceability query. Real-time information collection is achieved through Internet of Things devices, while data privacy is protected using encryption algorithms related to blockchain. Smart contracts are invoked to ensure secure storage and trusted traceability of data.

5.1. Traceability Data Acquisition

In the process of assembly and manufacturing, real-time monitoring equipment is utilized to monitor the production and processing of parts, ensuring the authenticity and credibility of data sources through real-time uploading by automated mechanical equipment. The assembly and manufacturing production line employs intelligent gateways to communicate with the production line PLC, enabling real-time acquisition of sensor-collected data and remote monitoring of operating statuses and process parameters for each equipment on the monitoring center platform. During logistics transportation, RFID, GPS, and other technologies are employed to track goods’ locations in real time, facilitating quick and accurate collection of logistics data as well as traceability of logistics information.

5.2. Traceable Data Storage Uplink

The process of storing traceability data in the blockchain is divided into three stages: pre-chain data processing, on-chain transaction broadcasting, and block consensus on-chain. The sub-assembly node M submits identity authentication information to the system and logs into the traceability system after successfully passing through both the traceability system and alliance chain access. Subsequently, it submits traceability uplink information to the system as depicted in the following steps.

Step 1: The sub-assembly node M uploads the original data I generated during the production process. The system calculates the hash value H(I) of I and then invokes the data uplink contract to transmit the unique trace ZID, hash value H(I), and trace key information $Path\_I$ to the blockchain server Q.

$$H(I)=H_{SHA256}(I)$$

(1)

$$M\to Q:\{ZID,H(I),Path\_I)$$

(2)

Step 2: Upon receiving the uplink request for traceability information, the blockchain server Q computes the hash values of ZID, H(I), and $Path\_I$. It then utilizes the private key d_M belonging to sub-assembly node M to sign h(M) using the ECDSA algorithm, thereby obtaining the digital signature $Sig\_I$ from the responsible person in charge of the link.

$$h(M)=H_{SHA-1}(ZID||h(M)||Path\_I)$$

(3)

$$Sig\_I=ECDSA(h(M),d_M)$$

(4)

Step 3: The blockchain server Q transmits the trace source ZID, H(I), $Path\_I$, and digital signature $Sig\_I$ to other nodes Ni within the blockchain system.

$$Q\to N_i:\{ZID,H(I),Path\_I,Sig\_I)$$

(5)

Step 4: The node Ni, upon receiving a message from the blockchain server Q, assembles the information into a transaction and disseminates it to other nodes within the network. Once more than half of the nodes in the entire network have reached an agreement on the transaction, consensus is achieved. Subsequently, the accounting node packages this validated transaction into a block and stores it on the chain.

After the successful storage of traceability data in the blockchain, the system performs data synchronization operations to retrieve transaction details and block unique identifiers stored in the latest block. These pieces of information are then stored in the local database to ensure seamless synchronization between on-chain and off-chain data. Simultaneously, appropriate access rights need to be assigned for different user identities and roles to guarantee secure storage of on-chain data and isolation of private data. Only authorized users can access specific traceable data. The complete winding process is illustrated in Figure 11.

5.3. Efficient Examination of Traceability Data

After purchasing a product, consumers can quickly access comprehensive information about the entire production-to-sales process, including parts sourcing, product parameters, and distribution channels. This is made possible through the unique identification of the product in the traceability system, enabling a more accurate assessment of product quality. The inspection of traceability information involves qualification verification, on-chain data query, and off-chain data verification. The specific steps are as follows.

(1)

The consumers log into the traceability system and provide personal information, such as their name and contact details, through the registration interface. They also submit relevant supporting documents required for traceability qualification certification, including the purchase order number, order creation time, transaction location, etc.

(2)

After receiving consumer support materials, the system manager verifies the authenticity of the information. Once the audit is successfully completed, a unique traceability access code is issued to the consumer, enabling them to enter the traceability data inspection module using this code.

(3)

The consumers input the unique traceability code ZID to inquire about product parameters, component origins, and logistics information. This includes crucial traceability data stored on the blockchain, such as hash values and transaction hash addresses. Additionally, they have the ability to file complaints regarding problematic products and manufacturers.

(4)

After receiving the user’s complaint, the supervisor will investigate the specific quality information of the batch of products produced by the enterprise in question based on the order number and specific content of the complaint. The backend system can retrieve comprehensive traceability data through the transaction address stored in the background database.

(5)

The regulator conducts an anti-tamper check on the complete traceability data within the blockchain, and verifies whether the hash summary stored in the blockchain aligns with the hash value derived from decrypting the digital signature and extracting complete traceability information from the local database. In case of inconsistency, it indicates potential data tampering, thereby determining that the credibility of enterprise’s locally stored quality traceability information is compromised, necessitating immediate initiation of responsibility investigation procedures.

The efficient and reliable traceability of the entire life cycle of the assembly and manufacturing supply chain can be achieved through the seamless integration of on-chain query and off-chain verification, enabling enterprises to establish a comprehensive and dependable inspection system for supply chain traceability data. Figure 12 illustrates the sequence diagram of the inspection process for traceability data in the assembly manufacturing supply chain.

6. Case Analysis

To verify the feasibility of the efficient traceability verification method proposed in this paper, this paper takes the supply chain traceability of a certain model of a complete vehicle manufactured by a certain complete vehicle manufacturing enterprise on a cloud platform as an example. From the perspective of the regulatory authority, the traceability information of the entire chain of the complete vehicle product associated with the unique order number submitted by the consumer is queried, and the reliability and credibility of the obtained traceability results are verified through specific cases.

6.1. Case Validation

During the traceability process, the consumer submits the vehicle transaction number, and the system randomly selects the transaction record with the order number “91535211710887” as the verification object. The traceability system retrieves the corresponding vehicle traceability information based on this order number, covering key data such as the unique vehicle identification code, technical parameters, production date, and quality inspection results. At the same time, it can trace the detailed technical parameters and manufacturer information of core assembly components (such as engines, transmissions, etc.). Additionally, the system supports the backtracking of upstream information such as parts suppliers, production dates, and batch numbers, ensuring that the source of parts is traceable and quality is controllable. All the above information is stored in the traceability blockchain and organized and presented in the form of block storage indexes. Some of the vehicle supply chain traceability query results are shown in

Table 3. Description of Vehicle Supply Chain Traceability Query Information.

Order Number	Transaction Manufacturer	Supply Chain Segment	Product Name	Blockchain Storage Index
91535211710887	Vehicle Manufacturer	Final Assembly	Vehicle TA-6000-B	35027396aa96cda4f6a7 59ce54d841afa73179b4992dcc6e
	Engine Manufacturer	Sub-Assembly	Engine WD6 15.95E	2f9b1fde9487a1f4820a91ad6e2e4 77eaf4ea16919784e951796f4fed4c67cec
	Connecting Rod Manufacturer	Component Level	Connecting Rod CA-6000-B	65dd3e29ba67f678b2bd3cobd37860732Ca26837b1d62a0106ab544e6e938761
	Camshaft Manufacturer	Component Level	Camshaft TA-6000-8	53fef53c032140279c8535a61c54ed 8cf6925770c4579fc3e1567cbb30ffcc12

.

Subsequently, the system queries the specific on-chain storage content of this information in the traceability blockchain through a peer-to-peer mechanism based on the block storage index. Taking the engine as an example, the relevant data includes detailed engine parameters, the hash value of privacy data, and digital signatures. The system decrypts the digital signature using the public key of the engine manufacturer and recalculates the hash value of the off-chain data to compare it with the original on-chain hash value, thereby determining whether the off-chain data has been tampered with.

As shown in the comparison results of engine data tampering verification in

Table 4. Comparison of Traceability Data Tampering Verification.

Data Type	Content	Processing Type	Processed Value
Original Data	PartName: “Connecting Rod CA-6000-B” PartID: “E4056983” AP_Quality: “Qualified”	Digital Signature	xNgw4i5g7jk0MugxOYiujARe/es 0R8uE+8TRWrPq5VMdrhf43isgXXcJ Pw3UrChXBEgLC8T0pLTex46vc0hc8g==
		Decrypted Hash	7008ce649c55b90adc3de5171f1c91 543ed80a368e28039de0055f4818722b2b
Tampered Data	PartName: “Connecting Rod CA-6000-B” PartID: “E4056980” AP_Quality: “Qualified”	Same Hash	2e0677896a1b4b3827bcb5d4b5e915 feff29326ab471f1176ca7492e4b97efdc

, the hash values obtained after processing the tampered data with the same hash algorithm are significantly different from the original on-chain hash values. Thus, it can be determined that the off-chain traceability data has been illegally modified, and its authenticity and credibility are invalid.

6.2. Performance Analysis of the Scheme

To verify the effectiveness of the efficient traceability verification method proposed in this chapter, this section compares the performance of the traditional Merkle index method and the efficient traceability verification method proposed in this paper in terms of time complexity and traceability query efficiency based on the Hyperledger Fabric blockchain platform (v2.4.9). The experimental environment is deployed on the Ubuntu 22.04 operating system, with a hardware configuration of an Intel Core i7-10300H CPU processor (Intel Corporation, Santa Clara, CA, USA), 20 GB of virtual machine memory, and Docker containers are used and orchestrated through Docker Compose (v1.24.1) for each node service.

(1)

Time complexity analysis

The traditional blockchain-based traceability query method usually starts from the latest generated block and traverses each block backward until the block containing the target traceability information is located. Assuming there are m blocks in the traceability chain and each block stores an average of n traceability data, the time complexity of this method is O(m × n), which incurs a relatively high retrieval cost in large-scale data scenarios.

In contrast, the efficient traceability verification method proposed in this chapter introduces a B-Merkle structure in the main chain and uses a Bloom filter to quickly determine the candidate block range to which the target traceability code belongs, effectively avoiding invalid searches of irrelevant blocks, reducing the complexity of main chain retrieval to O(log m); in the side chain, a skip list is introduced as a multi-level index structure to achieve rapid positioning of specific traceability content, with a time complexity of O(log n). Therefore, the overall time complexity of the traceability verification method is O(log m + log n), which is significantly better than the traditional linear traversal strategy.

Given the large number of participants in the assembly and manufacturing supply chain and the vast scale of traceability data, this method has a significant advantage in terms of time complexity. It can significantly shorten the response time of traceability queries while ensuring data integrity and security, thereby enhancing the overall operational efficiency and real-time response capability of the supply chain system.

(2)

Efficiency of traceability query

Under the condition that the block height remains unchanged, the test examines the performance difference in query between the traditional Merkle index method and the efficient traceability verification method proposed in this paper as the number of transactions stored in the block increases, with different scales of traceability data. The results are shown in Figure 13.

As shown in Figure 13, when the amount of traceability data stored in a single block is small, the query time difference between the two methods is relatively small; however, as the number of transactions within the block continues to increase, the gap in query time between the two becomes significantly larger. The reason for this is that the traditional Merkle index needs to traverse all transaction records to locate the target object, while the method proposed in this paper uses the feedback mechanism of the Bloom filter to perform effective pruning operations on the Merkle tree that does not contain the target information, quickly locking the hash value of the target block; at the same time, it combines a multi-level index structure to achieve parallel retrieval, further improving the efficiency of transaction location. Compared with the traditional Merkle index, the efficient traceability verification method proposed in this paper demonstrates superior query performance in large-scale data scenarios.

Under the condition that the number of traceability transactions stored in a block remains constant, the query performance of the traditional Merkle index and the method proposed in this paper was tested as the length of the blockchain (i.e., the total number of blocks) increases. The results are shown in Figure 14.

As shown in Figure 14, with the increase in the number of blocks on the traceability chain, the query time of both schemes approximately follows a linear growth trend. However, the efficient verification method proposed in this paper consistently outperforms the traditional Merkle index in terms of query response time. When the total number of blocks is 300, the average retrieval time of the traditional Merkle index is approximately 42.3 ms, while the proposed traceability verification method in this paper has an average time of 19.5 ms under the same conditions, achieving a query efficiency improvement of over 50%. This is mainly due to the fact that the method proposed in this paper can determine that a block does not contain the target information by judging the Bloom filter identifier value in the block header, and thus can skip the block without further traversing its internal nodes or performing previous transaction queries, effectively reducing unnecessary computational overhead and significantly lowering the overall query time.

7. Conclusions

To address the key issues existing in the traceability process of assembly manufacturing supply chains under the cloud model, this study proposes an innovative and trustworthy traceability solution. To verify the effectiveness of the proposed method, this paper selects the vehicle assembly supply chain as a typical application scenario to conduct empirical analysis. By comparing the time complexity and block retrieval efficiency, it can be seen that compared with the traditional block traceability retrieval method based on hash pointers and Merkle trees, the efficient traceability verification method proposed in this paper shows better performance in meeting the efficient retrieval requirements of traceability data in assembly manufacturing supply chains. Based on the current research results, the subsequent research will focus on the following two directions for in-depth exploration:

(1)

Supply chain traceability data analysis and resilience optimization. Relying on high-precision supply chain data obtained through a multi-chain traceability architecture, combined with key performance indicator (KPI) analysis and anomaly detection algorithms, identify bottleneck links and potential risk points in the supply chain network. Further integrate discrete event simulation (DES) and multi-agent simulation (MAS) methods to construct a dynamic supply chain optimization model, simulating the system response behavior under different disturbance scenarios such as raw material shortages, logistics disruptions, and demand fluctuations. Through sensitivity analysis and Pareto optimal solution, propose targeted supply chain resilience improvement strategies to achieve adaptive regulation capabilities for multi-agent collaborative assembly manufacturing supply chains.

(2)

Data-driven demand forecasting and inventory optimization. By leveraging deep learning techniques such as LSTM time series prediction models and graph neural networks, deep feature extraction and correlation relationship mining are conducted on the entire supply chain data. Through the integration of historical operational data, market intelligence, and macroeconomic indicators, an end-to-end dynamic demand forecasting model is constructed to achieve adaptive adjustment of safety stock levels. Under the constraint of service level, inventory holding costs and stockout losses can be effectively reduced.

The above-mentioned research will further improve the intelligent decision-making system of the assembly manufacturing supply chain under the cloud model, providing the industry with technical paths and solutions that are both theoretically innovative and practically valuable.