PaperlayerView exact revision on arXiv

Source-linked AI summary

Blockchain-based Digital Twins: Research Trends, Issues, and Future Challenges

Sabah Suhail, Rasheed Hussain, Raja Jurdak, Alma Oracevic, Khaled Salah, Raimundas Matulevičius, Choong Seon Hong

arXiv:2103.11585v2cs.CR

TL;DR

Industrial processes need trustworthy data for decisions, but disparate repositories and untrustworthy dissemination complicate its use. This survey reviews blockchain-based DT research, proposes a framework, and discusses research and deployment challenges. Its conclusion identifies shared-ledger data management and AI-supported predictive maintenance and threat intelligence as key outcomes.

  • Problem

    Industrial decision-making needs trustworthy data, while disparate repositories and untrustworthy data dissemination remain challenges for blockchain-supported DTs.

    Source: §1 · p. 2, §1.2 · p. 4

  • Method

    The survey systematically reviews blockchain-based DT design and implementation, proposes a trustworthy framework, and examines technical, logistical, and social challenges.

    Source: §1.2 · p. 4, §3.4.1 · p. 17, §5 · p. 21

  • Results

    The survey identifies a trustworthy 7D blockchain-based DT framework and concludes that blockchain supports shared-ledger data management while AI supports predictive maintenance and threat intelligence.

    Source: §3.4.1 · p. 17, §6 · p. 26

  • Takeaways & Limitations

    The authors conclude that research and deployment challenges need further investigation for successful implementation of blockchain-based DTs.

    Source: §5 · p. 21, §6 · p. 26

    Existing blockchain-based DT works overlook safety and security rules, with only one cited work recommending their adoption.

    Source: §3.1.3 · p. 11

Abstract

from arXiv · show

Industrial processes rely on sensory data for decision-making processes, risk assessment, and performance evaluation. Extracting actionable insights from the collected data calls for an infrastructure that can ensure the dissemination of trustworthy data. For the physical data to be trustworthy, it needs to be cross-validated through multiple sensor sources with overlapping fields of view. Cross-validated data can then be stored on the blockchain, to maintain its integrity and trustworthiness. Once trustworthy data is recorded on the blockchain, product lifecycle events can be fed into data-driven systems for process monitoring, diagnostics, and optimized control. In this regard, Digital Twins (DTs) can be leveraged to draw intelligent conclusions from data by identifying the faults and recommending precautionary measures ahead of critical events. Empowering DTs with blockchain in industrial use-cases targets key challenges of disparate data repositories, untrustworthy data dissemination, and the need for predictive maintenance. In this survey, while highlighting the key benefits of using blockchain-based DTs, we present a comprehensive review of the state-of-the-art research results for blockchain-based DTs. Based on the current research trends, we discuss a trustworthy blockchain-based DTs framework. We highlight the role of Artificial Intelligence (AI) in blockchain-based DTs. Furthermore, we discuss current and future research and deployment challenges of blockchain-supported DTs that require further investigation.

1 INTRODUCTION

Industrial decision-making depends on trustworthy sensor and lifecycle data, yet integrating IT and OT systems introduces attack vectors. The survey reviews blockchain-based digital twins (DTs), their research gaps, and a framework and challenges for industrial use.

1 INTRODUCTION

  • Integrating industrial IT and OT creates attack vectors, making trustworthy sensor data important for high-assurance decisions.Blockchain can support shared, secure product-lifecycle records, while DTs analyze data to help predict and optimize industrial operations.Source: §1 · pp. 2–3

1.1 Existing Works

  • Earlier surveys discuss blockchain and DT benefits, barriers, and applications but do not review the existing literature on blockchain-based DT approaches.Other work addresses lifecycle integration or focuses mainly on blockchain or DTs, rather than surveying blockchain-based DT approaches in particular.Source: §1.1 · p. 3

1.2 Scope of This Survey

  • The survey systematically reviews industrial blockchain-based DT design and implementation, proposes a trustworthy framework, and examines technical, logistical, and social challenges.It maps requirements from both blockchain and DT perspectives and discusses adoption approaches and open challenges.Source: §1.1 · p. 3, §1.2 · p. 4

2 BACKGROUND

The background introduces DTs and blockchain, then describes how their integration may support trustworthy data management across industrial applications and lifecycle participants.

2.1 Digital Twins

  • DTs map physical and virtual spaces through a connection; the 5D model adds DT data and services for managing, evolving, and optimizing both spaces.Industrial DTs use data to support predictions, decisions, fault diagnosis, and maintenance planning.Source: §2.1 · pp. 4–5

2.2 Blockchain

  • Blockchain records transactions in a shared ledger and links blocks with hashes, making tampering detectable; its industrial uses include supply chains and asset management.The survey frames blockchain-based DT use around secure data management and asks whether blockchain is appropriate given data readers and writers, trusted parties, identities, and verifiability.Source: §2.2 · pp. 5–6, §2.3 · p. 6

2.3.1 Trustworthy Sources; 2.3.2 Distributed Infrastructure

  • Blockchain can cross-validate sensor data from registered devices and provide distributed access to DT, production, supply-chain, and provenance data across lifecycle participants.DT data can also be shared among partially trusted parties with confidentiality, integrity, and availability protections.Source: §2.3.1 · pp. 6–7, §2.3.2 · p. 7, §2.3.3 · p. 7

2.3.3 Securing Digital Twins Data; 2.3.4 Data Traceability in Digital Twins; 2.3.5 Access Privileges of DTs

  • Blockchain provenance can trace why, where, and how data changed, while access controls restrict viewing and modification to authorized entities.The paper distinguishes a Digital Twin Instance (DTI), which mirrors an asset or activity, from a Digital Twin Aggregate (DTA), which combines DTIs and may expose a product line to manipulated updates.Source: §2.3.4 · p. 7, §2.3.5 · p. 7, §2.3.6 · p. 7

2.3.6 Safeguarding Product Lifecycle Data; 2.3.7 Automation Through Smart Contract

  • Smart contracts can automate application-dependent scenarios, while blockchain certificates and DT digitization are described as ways to establish product authenticity and identify counterfeit DT replicas.The text also introduces logging data as physical states change, but that passage does not complete its description.Source: §2.3.7 · pp. 7–8, §2.3.8 · p. 8

3 BLOCKCHAIN-BASED DIGITAL TWINS: RESEARCH TRENDS

The survey reviews blockchain-based DT design and implementation research, organizing current work through a taxonomy and comparison criteria. It reports growing research interest in integrating blockchain with DTs.

3 BLOCKCHAIN-BASED DIGITAL TWINS: RESEARCH TRENDS

  • The survey compares blockchain-based DT approaches across DT design, blockchain configuration, and security considerations.Its criteria include DT definition and composability, blockchain type and data storage, and security features such as encryption, access control, and quantum resilience.Source: §3 · p. 8, Table 1 · p. 9
  • The taxonomy organizes industrial blockchain-based DT design solutions, while the cited research trend figure indicates increased interest in integration.The section reviews design and implementation work and highlights challenges that remain unaddressed.Source: §3 · p. 8, Figure 4 · p. 8, Figure 5 · p. 10

3.1 Design Solutions for Digital Twins

The section identifies gaps in how DTs aggregate and correlate components, and reviews lifecycle, data integration, safety and security, and synchronization challenges. Existing work leaves open how to address these issues across a DT’s lifecycle.

3.1.1 Digital Twin Lifecycle.

  • Most lifecycle studies cover early or middle phases; dismissal remains underexplored, including how to preserve historical DT data for future production.Putz et al. mention dismissal but do not detail object disassembly or historical-data backup and access.Source: §3.1.1 · p. 11

3.1.2 Data Fusion and Interoperability.

  • Data fusion needs cleaning, format unification, and combination of heterogeneous sensor data, but surveyed works lack technical fusion details and practical implementation evidence.Interoperability is only partly addressed, and existing solutions do not adequately connect fusion to DT functionality.Source: §3.1.2 · p. 11

3.1.3 Safety and Security (S&S).

  • Safety and security rules are largely overlooked in DT design; Dietz et al. are identified as the sole work suggesting their adoption and monitoring.The section proposes exploring rule-modeling languages, security-aware device templates, behavior learning, and anomaly analysis.Source: §3.1.3 · p. 11

3.1.4 Synchronization and Consistency.

  • Sensor streams update from minutes to milliseconds, exceeding mainstream blockchain performance and leaving on-chain versus off-chain data selection unresolved.High throughput and low sharing latency are identified as requirements for synchronization between physical and virtual spaces.Source: §3.1.4 · p. 12
  • Existing work leaves data granularity and collection choices unresolved, including which data to sample and how many sensors to place and where.The section treats both data quality and quantity as important to precise decision-making.Source: §3.1.5 · p. 12

3.2 Design Solutions for Blockchain

The section surveys blockchain design choices and implementation gaps for DTs, from platform and contract choices to data storage, traceability, and trustworthy sources. It highlights limited practical evaluation and unresolved scalability and data-integrity challenges.

3.2.1 Implementation.

  • Only Putz et al. are reported to have completed a thorough implementation; most schemes are theoretical or partial, and practical industry evaluation remains limited.EtherTwin is an open-source prototype with performance and cost measurements; prototype suitability still needs validation by industrial experts.Source: §3.2.1 · pp. 12–13

3.2.2 Blockchain Type.

  • Most surveyed DT solutions use sequential public or private Ethereum or Hyperledger chains, while alternative blockchain structures may merit consideration for scalability.The cited examples use Ethereum [1] and Hyperledger; the section notes other DLTs may not yet be mature enough for mass adoption.Source: §3.2.2 · p. 13, §3.2.3 · p. 13

3.2.3 Scalability.

  • Smart contracts can automate DT functions including data sharing, twin creation, authorization, and event-based machine interaction [1].The section also describes possible uses for safety-rule triggers, PLC functions, simulation changes, and transaction auditing.Source: §3.2.4 · p. 13

3.2.4 Smart Contracts.

  • Consensus choice should reflect application requirements: PoW uses more computing power than PoA, while PoW suits partially trusted settings and PoA permissioned ones [1].XFT is used for higher throughput and lower latency; IBFT is suggested as a future prototype improvement.Source: §3.2.5 · p. 13

3.2.6 Data Storage.

  • Off-chain storage can address data volume and latency needs, but most proposals lack implementation details; Putz et al. implement storage using swarm DHT.The proposed pattern stores sensor feeds off-chain after registering sensors on-chain.Source: §3.2.6 · pp. 13–14
  • Off-chain storage may improve scalability and reduce on-chain costs, but introduces validity, centralization, real-time consistency, and security challenges.Source: §3.2.6 · p. 14

3.2.7 Traceability.

  • Traceability proposals often omit mechanisms or implementation; provenance containing source-destination pairs and key identifiers can be stored on-chain to reconstruct process chains [65].Source: §3.2.7 · p. 14

3.2.8 Trustworthy Sources.

  • Blockchain does not guarantee trustworthy data at its origin, so proposed safeguards include authorized-device registration, provenance and twinned-data checks, and participant inspections or penalties [64].The passage identifies device tampering, data forgery, and rogue participating entities as concerns.Source: §3.2.8 · p. 14

3.3 Design Solutions for Security

The section reviews encryption, access control, and quantum-resilience measures in blockchain-based DTs. It also introduces a seven-dimension DT framework and a high-level layered framework based on limitations in existing work.

3.3.1 Encryption.

  • Encryption proposals are often underspecified; Putz et al. implement AES-256 for data and encrypt symmetric keys with recipients’ public keys.Huang et al. propose encryption without implementation details.Source: §3.3.1 · p. 14

3.3.2 Access Control.

  • DT literature includes proposed role-permission mapping and a hybrid RBAC-ABAC model, but implementation details for access mechanisms are limited.Source: §3.3.2 · p. 15

3.3.3 Quantum Resilience.

  • Quantum computers may enable forged signatures and sabotaged transactions; surveyed responses include quantum-resistant schemes and a Twinchain variant, whose chosen hash-based signature lacks sufficient detail [66].The section also identifies quantum-linked communication and computing features as security research avenues.Source: §3.3.3 · p. 15
  • The authors present a seven-dimension DT framework and a high-level layered blockchain-based DT framework based on existing-work limitations.The frameworks connect DT, blockchain, and security design issues.Source: §3.3.3 · p. 16, Figure 8 · p. 16

3.4 Trustworthy Blockchain-based Digital Twins: A Unified Framework

The proposed framework combines blockchain-based data management with DT functions for trustworthy data analysis and feedback. Its design adds data integration and synchronization components, while addressing storage and security requirements.

3.4 Trustworthy Blockchain-based Digital Twins: A Unified Framework

  • The framework registers participants, maps physical and virtual data, stores data with provenance, builds models, and analyzes disturbances to trigger scheduling or model-calibration feedback [64].Its application layer analyzes data while the storage layer holds data and models.Source: §3.4 · p. 16

3.4.1 Design of Digital Twin.

  • The 7D model adds data wrangling, data fusion, and synchronization to the 5D DT model, and uses blockchain for secure data management; Table 2 maps model properties.The framework’s components are also mapped to autonomous-vehicle production as a CPS use case.Source: §3.4 · pp. 16–17, Table 2 · p. 17
  • Data wrangling cleans erroneous data and unifies formats, while fusion combines sources and composability integrates multiple DTs into an aggregate view [64].The framework also uses continuous physical-digital synchronization to check consistency and support self-adaptation.Source: §3.4.1 · pp. 17–18

3.4.2 Design of Blockchain.

  • Because sensor streams have high volume and velocity, the framework separates real-time sensor data flows from less dynamic production and provenance data rather than assuming all streams belong on-chain.The storage layer is described as using a lightweight, scalable, quantum-immune blockchain; multi-party access requires ownership-, role-, and level-based controls.Source: §3.4.2 · p. 18, §3.4.3 · p. 18

4 INTEGRATING BLOCKCHAIN-BASED DIGITAL TWINS WITH AI: WAY FORWARD

The section surveys how AI can support blockchain-based DTs through prediction, anomaly detection, and data-quality improvements, while identifying data-related challenges to adoption.

4 INTEGRATING BLOCKCHAIN-BASED DIGITAL TWINS WITH AI: WAY FORWARD

  • AI's predictive capabilities can help DTs analyze large data volumes, while blockchain and AI together are proposed for intelligent and trusted DTs.The section outlines AI's role and the outstanding challenges for AI-driven blockchain-based DTs.Source: §4 · pp. 18–19, Figure 9 · p. 19

4.1 How AI Will Set a Course for Blockchain-based Digital Twins

  • Interdependent industrial systems can face cascading effects from small variations, motivating analytics and threat intelligence to detect vulnerabilities, faulty models, and attack vectors.The section cites Stuxnet as an example of malicious code intercepting and modifying data sent to and from a PLC.Source: §4.1.1 · p. 19

4.1.1 AI-based anomaly detection solutions for DTs.

  • ML-based anomaly detection can use DT-generated training data and labeled anomalies from physical components to identify anomalies in industrial settings.The cited approach uses Siamese AutoEncoders tailored to weakly supervised learning with synthetic data.Source: §4.1.1 · p. 19
  • ML in DT synchronization and calibration can detect malicious or incorrect inputs, while ledger-stored safety rules and automated defenses can help contain disruptions.A provenance-aware blockchain can track entities that add or update the safety and security rules.Source: §4.1.1 · pp. 19–20
  • AI methods can improve data completeness and accuracy by imputing missing data, cleaning or augmenting data, and fusing heterogeneous sources.The section gives GANs as one method for imputing missing data and lists neural networks, Kalman filters, and Bayesian inference for data fusion.Source: §4.1.2 · p. 20

4.1.2 Improving Data Quality.

  • ML can project scenarios without direct observations, with algorithm choice depending on labeled-data availability; blockchain can record model parameters and prediction outcomes.The passage names LSTM, autoencoders, and reinforcement learning as possible approaches.Source: §4.1.3 · p. 20

4.2 Challenges for AI in Blockchain-based Digital Twins

  • The section identifies data-related challenges for AI in blockchain-based DTs, including model opacity, data scarcity, and biased or poor-quality inputs.It notes that these challenges must be addressed before AI-driven blockchain-based DTs are realized.Source: §4.2 · p. 20, §4.2.1 · p. 20, §4.2.2 · p. 21

4.2.1 Need for Explainable AI.

  • Explainable AI can make model decision processes more interpretable, while blockchain can track AI data flows and behaviors to support transparency.The section notes that opacity can hinder adoption, especially in safety-critical settings where errors may affect physical decisions.Source: §4.2.1 · p. 20

5 RESEARCH AND DEPLOYMENT CHALLENGES

The section frames technical, logistical, and social challenges as open issues for blockchain-supported DTs in industry.

5 RESEARCH AND DEPLOYMENT CHALLENGES

  • Technical, logistical, and social challenges remain open research and deployment issues for blockchain-based DTs in industry.The section says Fig. 10 presents a challenge taxonomy and Table 3 summarizes challenges, causes, and possible solutions.Source: §5 · p. 21
  • The section presents these challenges as avenues for future research into implementing blockchain-supported DTs in industry.It points to Fig. 10 for the taxonomy of current and future research and deployment challenges.Source: §5 · p. 21
  • The challenges are summarized alongside their causes and possible solutions in Table 3.Fig. 10 presents the detailed challenge taxonomy.Source: §5 · p. 21

5.1 Technical Issues

The section identifies technical challenges in representing DTs faithfully and managing the volume, timing, storage, and variety of industrial data.

5.1.1 Representation of Digital Twins.

  • A DT's high-fidelity model must mirror its physical counterpart's behavior and functions in real time, despite uncertainty and change in the underlying system.The section calls for cognitive DT solutions to address unknowns and accommodate model evolution as the physical object changes.Source: §5.1.1 · pp. 21–22
  • Replicas require synchronization with the physical object; blockchain transaction management is one proposed way to address consistency across replicas.Replication also raises issues of temporal dependencies, advance knowledge of events and activities, and the need for experts to solve complex problems.Source: §5.1.1 · p. 22

5.1.2 Data-related Challenges.

  • Data collection requires balancing volume and sampling frequency against prediction accuracy, information overload, transmission bottlenecks, and storage costs.The section recommends defining data's spatio-temporal granularity according to system requirements.Source: §5.1.2 · p. 22
  • High-volume blockchain storage and heterogeneous data remain challenges; off-chain storage, hashes, and removable blockchains are listed as alternatives.Different formats and sampling rates can obscure faults or lose details, and the section says big-data storage challenges remain open.Source: §5.1.2 · p. 23

5.2 Logistical Issues

The section describes financial, standardization, interoperability, and trust concerns that complicate industrial adoption of blockchain-based DTs.

5.2 Logistical Issues

  • The costs of connectivity, computing, bandwidth, storage, blockchain deployment, and specialist training can make the transition more beneficial for medium-to-large industries than small ones.The section also notes that smaller industries may eventually need to adopt these technologies as they scale or face technological obsolescence.Source: §5.2.1 · p. 23

5.2.1 Expenditures on Infrastructure.

  • Missing laws, shared data and model standards, and business-process integration hinder an internationally operational DT framework.The passage calls for interoperability across workflow stages and a universal, non-proprietary platform with user-friendly, API-based interfaces.Source: §5.2.2 · p. 23

5.2.2 Standardization Barriers and Efforts.

  • Blockchain platforms differ in data structures, scalability, fault tolerance, decentralization, and consensus, while non-interoperability fragments industrial ecosystems.The section describes standardization efforts by EEA, ISO, ITU-T, W3C, and SWIFT.Source: §5.2.2 · p. 24

5.2.3 Skepticism in Blockchain Technology.

  • Industrial adoption concerns include sensor reliability, latency, privacy, energy use, throughput, standardization, and legacy-system migration.The section also notes that blockchain's black-box nature can require consumers to trust processes without understanding their technical underpinnings, contributing to company hesitation.Source: §5.2.3 · p. 24

5.3 Social Issues

Blockchain-based digital twins face social, ethical, legal, and environmental challenges, including restricted sharing, unresolved rights and consent questions, and energy-intensive mining.

5.3 Social Issues

  • Industry rivalry and commercial sensitivity restrict data and model sharing, while concentration of DT ownership may disadvantage smaller firms and limit access.The passage calls for collaborative data governance and wider sharing of data and models.Source: §5.3.1 · p. 24

5.3.1 Ethical and Legal Issues.

  • Ownership rights for retired workers’ twins and rules for consent and anonymity in healthcare data remain unresolved, motivating privacy-preserving but transparent blockchain designs.Source: §5.3.1 · p. 25

5.3.2 Environmental Impact.

  • Compute-intensive mining makes blockchain consensus energy-intensive; proposed responses include more efficient or miner-free mechanisms and AI-supported energy planning.Ethereum’s estimated energy expenditure is almost 10 TWh.Source: §5.3.2 · p. 25

6 DISCUSSION AND CONCLUSION

Industrial digital twins use heterogeneous data to analyze, predict, and optimize operations, while blockchain securely records lifecycle events. The survey reviews industrial blockchain-based DT designs, proposes a 7D framework, and identifies AI roles and further research and deployment challenges.

6 DISCUSSION AND CONCLUSION

  • Digital twins analyze, predict, and optimize operations before real-world implementation, while blockchain securely records product lifecycle events and supports secure DT frameworks.Source: §6 · p. 26
  • Blockchain lets lifecycle stakeholders manage data on a shared ledger for trust, integrity, and security; comparison of existing work identified design gaps and informed a trustworthy 7D framework.Source: §6 · p. 26
  • AI integration in blockchain-based DTs supports intelligent twins for predictive maintenance and threat intelligence.Source: §6 · p. 26
  • The authors identify further research and deployment challenges as requiring investigation for successful blockchain-based DT implementation.Source: §6 · p. 26
  • The authors conjecture that the survey may help researchers address challenges and advance standardization of blockchain-based DTs in industrial applications.Source: §6 · p. 26
Loading 2103.11585v2…