Secure sharing of big digital twin data for smart manufacturing based on blockchain

Highlights

• The concept of Big Digital Twin Data (BDTD) is proposed.

• Blockchain and Cloud technology are applied to BDTD sharing.

• The implementation method is presented to support the timely sharing of BDTD.

• The time latency of BDTD sharing is evaluated to demonstrate feasibility.

• An algorithm for optimal sampling rate selection of BDTD is proposed.

Abstract

With the rapid development of digital twin technology, a large amount of digital twin data named as big digital twin data (BDTD), is generated in the lifecycle of equipment, which is supposed to be used in digital twin enabled applications. However, in the implementation of these applications, data sharing problem which is caused by the lack of data security as well as trust among stakeholders of equipment, limits data using value. It is a novel way to introduce blockchain technology into digital twin to solve the problem. However, current methods cannot fulfill the requirements of exponential growth and timely sharing of BDTD. Therefore, a blockchain-based framework for secure sharing of BDTD is proposed to solve the problems. Cloud storage is integrated into the framework, with which, BDTD is encrypted and stored in Cloud, while the hash of BDTD and transaction records are stored in blockchain. Some rules of generating new block are designed to improve the processing speed of blockchain. An algorithm for optimal sampling rate selection is presented to maximize total social benefits of the participants of BDTD sharing. Simulation results show that the algorithm is better than traditional method for maximizing the total social benefits. Furthermore, a protype system is developed and evaluated based on Fabric test network. Evaluation results show that BDTD can be shared securely multiple times per second through the framework, which demonstrates the feasibility of the framework in supporting timely sharing of BDTD.

Introduction

As one of the main developing trends of Industry 4.0, digital twin (DT) technology increasingly gains the attention of both academia and industry. As the precise virtual copy of product, digital twin almost mirrors every phases of product, such as design, manufacture, maintaining, and service [1]. The quality, cost, efficiency and life of product can be monitored, controlled, optimized, and predicted precisely by using DT with intelligent algorithms respectively [2]. In these applications, DT data is used to drive the self-updating of DT, train intelligent algorithms, and mine professional knowledge. In order to achieve high precision and high reliability of DT-based applications, DT data is collected from different equipment, different production line and even different plant. Furthermore, DT data usually has big volume, in order to meet the requirement of model training. Therefore, DT data sharing is required among different equipment, different lifecycle stages of equipment, different applications of equipment, and even different owners of equipment. On the one hand, because it is difficult for one equipment to gather enough data only by itself for DT-based application. On the other hand, DT data needs to be shared to promote lifecycle integration of equipment. For instance, DT data collected from using stage can be used to improve the design quality in design stage and make better manufacturing process in manufacturing stage. As shown in Fig. 1, the typical scenarios of DT data sharing include:

• Among different equipment with the same type

Because of the similarity of data in the same type of equipment, DT data can be shared to make DT-based services more precise. Data which is reusable and valuable for other equipment, needs to be shared. Such as the operation data before and after the failure of same type of bearings needs to be shared. On the contrary, business privacy data, useless data, and redundant data don't need to be shared. Such as the detailed normal operation data of bearing is redundant [3], and doesn’t need to be shared completely. Only the data during a period of time before and after bearing failure is valuable and needs to be shared.

• Among different lifecycle stages of equipment

There is time series relationship among DT data from different lifecycle stages of equipment. What’s more, the ownership of equipment may change in different lifecycle stages (e.g. design stage, manufacturing stage, integration stage, using stage, recycling stage). It causes that data islands are formed between different lifecycle stages of equipment [4]. Therefore, DT data needs to be shared among all lifecycle of equipment to promote lifecycle integration of equipment.

• Among different applications of equipment

The DT data from different applications, such as the design optimization, the operation optimization, and the prognostics health management (PHM) based on DT of equipment, is useful to each other. Hence DT data can be shared among different applications of equipment to improve the service quality of DT-based applications.

• Among different owners of equipment

Due to privacy and security, there are data islands between different owners of equipment. In order to speed up DT data aggregation, DT data of different owners’ equipment needs to be shared to fulfill the data requirements of DT-based applications.

With machine tool as an example, physical equipment data includes spindle speed, cutting tool vibration, temperature, and other various sensor data. These data are generated quickly and transferred to virtual space for updating virtual model or data analysis. Virtual equipment data refers to the simulation results and the results of data analysis. Such as simulation of gear wear based on virtual model, predictions of cutting tool life based on real-time data and huge amounts of historical data. Integrating the physical equipment data, virtual equipment data of machine tools as well as service system such as Manufacturing Execution System (MES) and Enterprise Resource Planning (ERP), accurate production scheduling [5,6], efficient flexible production [7], and other higher value applications [8] can be realized. Therefore, the shared DT data has big volume, many varieties, fast generation velocity, and great value, which are named as 4 Vs characteristics of big data. In other words, the DT data gathered from various sources is a kind of big data [9]. Big data can also be regarded as an important part of digital twin. In order to describe the DT data gathered from different equipment, different lifecycle stages of equipment, different applications, and different owners of equipment, big digital twin data (BDTD) is proposed to describe the above characteristics.

Although many DT-based applications have been studied, it can be observed that, for BDTD, as the impetus of DT-based applications, there is lack of a method to make BDTD be shared securely. Security is crucial for interested parties which want to share their data with trustless others. If the shared data is leaked, it might cause huge losses for the party. In addition, because of the lack of trust, the operations of payment and sending data involved in BDTD sharing cannot be implemented safely. Therefore, a secure method for BDTD sharing is required to promote BDTD sharing among trustless parties.

The first possible method is to design access control mechanism which relies on a third party. For instance, a service-oriented security architecture for access control in semantic data federations was presented to solve the problem of satisfying the confidentiality requirement of various stakeholders in product lifecycle [10]. A pull mode industrial solution was designed to enable data sharing between original equipment manufacturers and their suppliers by using a combination of PTC PDMLink and Microsoft SharePoint technologies [11].

The second probable method is to depend on Cloud/Fog computing. For instance, a secure online/offline data sharing framework was proposed to solve the problems of online/offline encryption, outsourced decryption, and fine-grained keyword search in Cloud [12]. A fog-computing-based approach was proposed to share industrial big data with high security by moving the integration task from the Cloud to the edge of networks [13].

However, the two methods both depend on centralized third party that is vulnerable to malicious attacks, such as single node invalidation and data tampering. It lacks a secure and decentralized method for data sharing among all parties in lifecycle of equipment.

In order to solve the problem, blockchain is introduced into industrial filed to improve the security of data sharing. For instance, a blockchain-based distributed peer to peer (P2P) network architecture was proposed to solve the problem of centralization in Cloud manufacturing [14]. Blockchain-based data sharing techniques were proposed to address the security problems of data transmission in Industrial Internet of Things (IIoT) [15,16]. A blockchain-enabled efficient data collection and secure sharing scheme combining Ethereum and deep reinforcement learning was proposed to ensure security when sharing data among smart mobile terminals [17]. Blockchain-based architectures combining RFID and IoT technology were proposed to provide a chain of immutable transactions of supply chains in multi-company project environments [18,19]. An industrial blockchain-based framework for product lifecycle management was proposed to fulfill the requirements of the openness, interoperability and decentralization [20]. Moreover, in order to support the application of CPS and DT, blockchain-based resource sharing architectures were proposed to improve the security and efficiency of data management [21]. The representative latest published researches related with blockchain and DT are compared as shown in Table 1.

As described in Table 1, Huang et al. proposed a blockchain-based method for data management of digital twin of product [25]. Zhang et al. proposed a blockchain-based architecture for the configuration of intelligent manufacturing systems [24]. Suhail et al. discussed the blockchain-based framework for the data management and security of Industrial Internet of Things (IIoT) [23]. Putz et al. proposed a blockchain-based sharing model for the management of digital twin components and associated information [22].

To the data sharing problems of DT, the time latency of data sharing is a very important factor. However, in the researches listed in Table 1, Putz et al. fully implemented their method which is based on Ethereum and Swarm, and evaluated the latency and cost for interaction [22]. Because Swarm is restricted to one update per second, timely DT data sharing, which need to support sensor data to be updated several times per second, might not have been considered in their research. In addition, Zhang et al. evaluated the throughput and latency of Ethereum and Fabric, which showed that Fabric could achieve higher throughput and lower latency than Ethereum [24]. But the time latency of the combination of Fabric and Cloud for data sharing has not been evaluated. If blockchain technology and Cloud technology are applied to DT data sharing, the evaluation of time latency is indispensable.

To this end, our goal is to design a blockchain-based framework which can support the secure sharing of BDTD, fulfill the processing speed requirement of time-sensitive data, and maximize the total social benefits of BDTD sellers and BDTD buyers. The main contributions of this paper are as follows.

• A framework for secure sharing of BDTD is proposed by combining blockchain with Cloud technology. Channel is used to achieve business isolation and data confidentiality for the participants of BDTD sharing. The hash of BDTD and transaction records are stored in blockchain. But original BDTD are encrypted, encapsulated, and stored in Cloud.

• The implementation method of the framework is presented. In order to fulfill the requirement of time-sensitive data, some rules of generating new block are designed to improve the processing speed of blockchain. Evaluation results demonstrate that BDTD can be shared securely multiple times per second through our framework.

• An algorithm for optimal sampling rate selection of BDTD is proposed to maximize the total social benefits of BDTD sellers and BDTD buyers. Simulation results show that our algorithm has better performance than traditional method in terms of maximizing the total social benefits.

The rest of this paper is organized as follows. Requirements for secure sharing of BDTD are analyzed in Section 2. A framework of blockchain enabled secure sharing of BDTD is proposed in Section 3. Section 4 describes the implementation method of secure sharing of BDTD based on blockchain. An algorithm for optimal sampling rate selection of BDTD is presented in Section 5. The framework and algorithm are evaluated in Section 6. Section 7 is discussion. Finally, conclusions are summarized in Section 8.