The Road to Accountable andDependable Manufacturing
Jan PennekampCommunication and Distributed Systems, RWTH Aachen University, Aachen, Germany
Roman MatzuttCommunication and Distributed Systems, RWTH Aachen University, Aachen, Germany
Salil S. KanhereSchool of Computer Science and Engineering, University of New South Wales, Sydney, Australia
Jens HillerCommunication and Distributed Systems, RWTH Aachen University, Aachen, Germany
Klaus WehrleCommunication and Distributed Systems, RWTH Aachen University, Aachen, Germany
Abstract—In manufacturing, advances from the IoT foster the vision of a highly dynamic andinterconnected Industrial IoT. However, business-driven use cases mandate different levels ofsecurity, privacy, accountability, and verifiability alike. Blockchain technology addresses theserequirements and thereby enables previously unforeseen collaborations. The authors emphasizethe need for active research at the intersection of IoT, CPS, and blockchain.
Authors’ version of a manuscript that was submitted for publication in Computer.
MANUFACTURING is expected to signif-icantly benefit from recent advances in the areasof Internet of Things (IoT) and Cyber-PhysicalSystems (CPS). Particular development directionsinclude establishing highly-dynamic business re-lations and creating interconnected productionenvironments, even for short-lived collaborations,through increasing degrees of automation basedon (sensor) data [1]. Concepts of the IndustrialIoT (IIoT) or Internet of Production (IoP) [2] ex-plicitly target to implement these improvements.
Research mainly evolves around three existingpillars (P0–P2): (P0) CPS and site-related im-provements (œ) with limited external influences,(P1) extended data sharing along the supplychain (Õ), e.g., to reduce the bullwhip effect,and (P2) secure industrial collaborations acrosssupply chains (Ö), e.g., to reduce ramp-up costs.To achieve P1 and P2 not only with today’s(established) long-term trust but also in settings
with dynamically evolving and flexible short-termrelationships, we identify a new research pillar(P3) that enables accountable and dependabledataflows for stakeholders without any trustedor previous relationships (v). In this article, wefocus on the research pillars P1–P3 that considermultiple stakeholders in collaborative processes.
Such industry-driven settings mandate specialneeds that traditional solutions in the IoT can-not satisfy. These aspects encompass improvedaccountability and verifiability to deal with un-certainty concerning the origin [3] and reliabilityof data [4], but also security and privacy re-quirements have to be considered as informationleakage can have tremendous consequences inhighly competitive environments [2]. We envi-sion that the consequent integration of blockchaintechnology provides these desired features by de-sign. Its tamperproofness offers verifiability andreliability once information has been recorded on
the blockchain. Similarly, blockchains are decen-tralized and thus well-suited for securing interac-tions among mutually distrustful parties. Finally,the extensible nature of blockchain technologyenables scalability features, such as sidechainsor sharding [5], as needed for solutions acrossdifferent use cases and domains.
Given that research at the intersection of IIoTand blockchain is still in its infancy, we identifythree key research areas. We discuss blockchain-specific research questions for the industrial set-ting, which mainly evolve around the generalscalability of proposed solutions and the privacyof participants. Similarly, we identify a lackof manufacturing-specific solutions that integrateblockchains to improve accountability in this do-main. We discuss scenario-driven research direc-tions that close this gap and realize fast, versa-tile, accountable, and dependable manufacturingenabled by blockchains. Furthermore, we discussarising socio-economic challenges. Particularly,new legal frameworks will need to take intoaccount the increased usage of external data, po-tentially in safety-critical applications. First andforemost, however, we want to raise awarenesson how to establish trust into the authenticityand correctness of data on the blockchain asa foundation for interorganizational data sharingwithin the IIoT.
MOTIVATION & POTENTIALS
Manufacturing is expected to compile vastamounts of process and product data in the nearfuture [2]. Consequentially, we have to deal withassociated big data challenges that stand out dueto virtually infinite volumes of available sensordata and the increased need for high-frequencysensing [1]. However, big data also providesopportunities when properly extracting its encap-sulated knowledge [1]. Regarding manufacturing,this potential has previously been neglected forlack of globally available process information,and even data sharing along the supply chainwas limited. Figure 1 illustrates the data sharingalong (Õ) and across (Ö) supply chains, whichwe detail hereafter based on two fine blankinglines.
Information Sharing along Supply Chains (P1Õ)
Traditionally, supply chain data sharing wasdriven by large companies dictating their require-ments to all suppliers. In this setting, informationwas collected in data sinks accessible by single(large) players [4], e.g., automotive manufactur-ers. Furthermore, due to privacy concerns, datais usually shielded from external stakeholders,for example, even rather insensitive information,such as delivery schedules or shipment tracking,is retained locally. Today, additional data is onlyshared under the promise of large financial im-pacts despite production data being expected toimprove manufacturers’ productivity and overallproduct quality [2].
This situation is unsatisfactory as it fails toaddress several desired aspects. Especially re-garding legislation, today’s landscape cannot reli-ably provide (long-term) verifiability of relevantinformation [6], e.g., provenance data for partsin the aerospace industry or associated mainte-nance protocols. Although additional processesare often in place, counterfeit or non-fair tradeproducts are, occasionally, still entering legitimatesupply chains [7]. To improve the reliability of(received) data, we envision technical solutionsthat minimize the room for manipulations andprovide an efficiently verifiable certification foreach individual product. Furthermore, a unifiedapproach could improve governmental oversight,which is especially desirable for safety-criticalproducts or food chains [8].
Another insufficiency stems from the lackingidentifiability of root causes of manufacturingor product failures [6]. Currently, accountabilityis mostly limited to contractually-bound stake-holders. If not explicitly contractually negotiated,individual untrusted suppliers may remain pas-sive or even behave adversely for their benefits,e.g., when covering up incidents. Simultaneously,missing feedback to estimate the lifetime or thefit of a product, which both might depend onthe application, hinders the implementation ofimprovements. To overcome such limits, acces-sible production and usage data can provide in-sights [2].
2 Computer
Supply Chain Flow
Dataflow (P2 ⇅)
Aerospace Fine Blanking Manufacturer
Automotive Fine Blanking Manufacturer
Lubricant Supplier
Material Supplier
Automotive Assembly
Tool Manufacturer Aerospace Assembly
Dataflow (P1 ⇄)
Figure 1: Manufacturing engulfs both dataflows along the supply chain (P1 Õ)and across supplychains (P2 Ö). Suppliers (here: for lubricants, material, and tools) support manufacturers whothemselves provide subsequent assembly lines with production data. Similarly, manufacturers exchangeprocess information (here: fine blanking lines), the processed material, and their interplay. A currentlynon-existing relationship between both assembling companies could be non-existent due to theuntrusted environment (P3 v). We adapted the figure from our analysis of dataflows in an Internet ofProduction [2].
Foundations for Expanded Secure IndustrialCollaboration Across Supply Chains (P2 Ö)
In addition to the marginal data sharing alongsupply chains (P1 Õ), data exchanges acrosssupply chains (P2 Ö)are basically non-existingin today’s manufacturing landscape [1]. Whilemanufacturers gather usage data from their cus-tomers (in centralized data silos), virtually noknowledge exchange happens between differentoperators of (identical) machines [2]. For ex-ample, experiences with used machine config-urations or information about the (expectable)production quality can reveal interesting insightsinto newly configured manufacturing processes.Hence, all knowledge is retained locally withoutglobal availability, despite potentially tremendousbenefits [2].
To improve productivity and to decrease costs,companies could, for instance, share ideal ma-chine configurations for their workpieces, e.g.,within their fine blanking line, without revealingall details to the machine supplier. Furthermore,this information exchange may reduce ramp-uptimes of new manufacturing processes by deriv-ing machine parameters from readily availableinformation (cf. Figure 1). Consequentially, non-competing companies can cooperate and jointlyassemble a shared knowledge base in a give-and-take manner or offer their valuable data for sale.As of today, a lot of expected potential is still
unexplored.
Ad-Hoc Relationships in UntrustedEnvironments (P3 v)
When considering relationships with previ-ously unaffiliated and thus untrusted compa-nies (P3 v), several additional use cases emerge.Along supply chains (P1 Õ), identifying theideal supplier for a component is simplified whenthe utilization of relationships among previouslyunaffiliated parties is improved. Similarly, ex-changing information with companies in relateddomains across supply chains (P2 Ö)is currentlyhindered by a lack of trust between the in-volved stakeholders. We expect that more usecases surface once the first steps towards secureindustrial collaboration have been taken as busi-nesses are naturally cautious when sharing sen-sitive and valuable details, especially productionand product data [2]. Furthermore, we observethat currently no uniform standardization for datasharing exists, which especially hinders flexiblerelationships as company-specific adjustments arerequired for each new partner [4].
In the context of accountable and dependablemanufacturing, we also have to address privacyand safety [9]. Appropriate means are not yetavailable, or they are not proven or tested inmanufacturing [1]. A major milestone to establishtrust can be achieved by providing accountability,
September 2020 3
Feature
verifiability, and transparency for all actions andtraded information. Consequentially, blockchainsare a promising tool to establish trust in mutuallydistrustful manufacturing markets and to eventu-ally allow for interorganizational data sharing andnovel applications.
THE INFLUENCE OF BLOCKCHAINSBlockchain systems have matured consider-
ably since their introduction through Bitcoin in2008. Initially created for the decentralized, yetsecure, management of digital currency, the po-tential of blockchains for larger and more diversetasks was quickly identified across academia andindustry.
The State of Blockchain IntegrationWe now reiterate impactful milestones and
applications of distributed ledger technology toassess its current level of integration into businessprocesses and to identify areas where blockchainshave been applied successfully.
Financial Origins Bitcoin paved the way forglobal financial transactions without banks asintermediaries. Besides inspiring numerous com-parable cryptocurrencies, the banking sector alsonoticed the potential of blockchains to improvetransactions between financial institutes. This de-velopment yielded major blockchain-based inter-bank networks, e.g., the Ripple payment andexchange network or JP Morgan’s InterbankInformation Network. Furthermore, blockchainspromise to provide better, i.e., more direct, cus-tomer experience at lower costs due to more auto-mated, disintermediated processes. Especially inscenarios where participants are known, and theirmajority is trusted, consortium blockchains areseen as key enablers for shaping new transactionprocesses in highly distributed applications, e.g.,accounting in supply chains.
Digital Assets One of the first non-cryptocurrency applications of blockchainswas the establishment of digital assets and notaryservices. While dedicated solutions, such asNamecoin, were launched quite early, numeroussuch services piggyback on existing blockchains,commonly Bitcoin [10]. Particularly, to transfer
digital ownership of property, coupons, or stock-marketing shares through a cryptocurrency’sblockchain, users can tie assets to blockchaintransactions. Beyond that, notary servicesimmutably attest the existence of documents bystoring a cryptographic hash on a blockchain,a tamperproof identifier to which owners cansubsequently refer to.
Process Automation Smart contracts [5] re-alize the automated execution of transactionsonce the blockchain’s state satisfies their one-timeprogrammable conditions. This tamperproof pro-grammability allows for transparent automationof global processes. While Ethereum popular-ized blockchain-based smart contracts, businessapplications are commonly built using consor-tium blockchains, e.g., created through Hyper-ledger Fabric or the Ethereum-compatible Quo-rum. Beyond the banking sector, insurers pro-cess insurance claims without human interactionthrough smart contracts. An increased demandfor blockchain-based process automation sparkedthe creation of Blockchain-as-a-Service solutions,e.g., offered by Microsoft Azure, IBM, and Ama-zon Web Services. These services lower the bar-rier for creating blockchain-backed architectures,but also introduce an infrastructure provider as anew centralized entity.
Internet of Things Advances in process au-tomation proliferated the vision of coupling au-tonomous IoT devices with blockchains. Themain advantages of blockchain-based IoT infras-tructures lie in the immutable and decentralizedIoT-based sensing of physical environments inconjunction with the accountable recording ofactuation events. If seized well, these capabilitiescan significantly simplify applications for smartcities, e.g., smart microgrids [11] or vehicularnetworks [12]. Here, blockchains aid trust man-agement and access control to sensed data alike.
Supply Chain Blockchains may be used asan architectural pillar for reshaping supplychains [13], [7], [6], [14], especially due toimproved financial transactions, asset manage-ment, process automation, and data manage-ment. However, smooth integration is still lack-ing [9]. TrustChain [8] or ProductChain [3] al-
4 Computer
ready tackle important issues of supply chaindeployments, such as reputation-based trust man-agement among suppliers and provenance track-ing for customers. Still, holistic, all-encompassingapproaches to improve supply chains based ondistributed ledgers are yet to come.
Useful Properties for Diverse Applications
Even today’s limited integration of blockchaintechnology into business processes highlights thatdistributed ledgers have proved to provide valu-able foundations for various domains, applica-tions, and use cases. Particularly, we highlightthat blockchain technology provides desirablecontributions to flexible collaborations and es-pecially to applications involving supply chains.First, the decentralized nature of blockchain ap-plications suits the highly distributed and het-erogeneous environments created by collabo-rating companies and supply chains. Second,blockchains can provide data integrity and verifia-bility even if collaborators are partially distrustingeach other. As part of this process, recorded datais kept on a tamperproof ledger. Finally, estab-lished measures to keep track of digital assetsand to prevent double-spending enable the pub-lic, transparent traceability of products or theircomponents. However, the decentralization andimmutability of blockchains creates issues thatwere not present in traditional business processes.Next, we thus dive into resulting challenges that,once tackled, will help realize suitable full-stacksolutions for improving business processes viadistributed ledgers.
OPEN RESEARCH AREASWe identify three layers of open research
areas that we illustrate in Figure 2: (L1) yetunaddressed challenges for the use of blockchaintechnology in manufacturing, (L2) new opportu-nities for a fast, versatile, accountable, and de-pendable manufacturing enabled by blockchains,i.e., scenario-driven challenges, and (L3) socio-economic challenges stemming from immutablyrecorded production data and highly flexiblecross-company collaborations. We consider theselayers to be highly relevant when shaping thefuture of interconnected manufacturing.
Open Blockchain-Inherent Challenges (L1)As groundwork for more scenario-specific re-
search, we identify blockchain-induced researchareas that surface when relying on blockchainsfor accountable and dependable manufacturing.
Scalability Permissionless blockchains tradi-tionally struggle with limited scalability in termsof transaction throughput, transaction latency, andstorage requirements. For instance, Bitcoin fa-mously has a low transaction rate of only 3.5transactions per second as its 10-minute inter-block delay requires users to wait for an hourto safely accept payments [5]. Even though con-sortium blockchains can utilize more efficientconsensus algorithms [15], recording large num-bers of events on-chain still remains challeng-ing. Solutions may aggregate multiple eventsinto single or few (on-chain) transactions, similarto micropayment channels that boost transactionthroughputs in today’s cryptocurrencies. Further-more, applying sharding schemes [5] to consor-tium blockchains may improve their transactionthroughput as these schemes target to partitionthe network and to distribute the responsibilityfor transaction processing.
Another scalability issue is the ever-increasingstorage requirement to operate blockchains. Forinstance, heavily-utilized blockchains today ac-cumulate hundreds of Gigabytes of historicaldata. This problem is aggravated in the contextof supply chain applications once suppliers arerequired to tie their reports for other contrac-tors immutably to the blockchain. Pruning strate-gies have been proposed to unburden blockchainnodes from storing historic transaction data thathas become obsolete meanwhile [16]. However,applications relying on blockchain-extrinsic datacannot immediately seize this potential since whatconstitutes obsolete data has to be defined ona per-application basis. Again, also partitioningdata storage across the network with shardingschemes can reduce per-node storage require-ments. Overall, future research needs to assessthe need for long-term data availability to allowfor efficient and scalable solutions.
Efficiency Wide-spread adoption ofblockchain technology in supply chainsnecessitates an efficient operation of the
September 2020 5
Feature
Scen
ario
-Driv
enC
halle
nges
Soci
o-Ec
onom
icC
halle
nges
Blo
ckch
ain-
Inhe
rent
Cha
lleng
esLegal Frameworks
(Governmental Oversight)Access & Transparency
(Platform Openness)
ReliableProduct Information
Efficient & DependableCollaboration
DynamicDistributed Markets
Scalability Efficiency Immutability Privacy
Blockchain size
Transaction throughput
Distributing responsibility
Operational costs
Workload on nodes
Environmental impact
Persisted garbage
Outdated data
Correctness of information
Sensitive metadata
Information leakage
Verifiability & transparency
• Accountability along the supply chain
• Correctness of available information
• Tamperproofness of measurements
• Untampered digital processing• Linking of physical goods and its data
• Granularity of data sharing
• Private accountable billing of companies
• Keeping automation with more flexibility
• Granting access to external companies• Dynamic digital factories
• Trade-off privacy vs. verifiability
• Fairness of data sharing
• Maintaining a data catalogue
• Privacy-preserving bidding platform• Measuring the value of data
Research
Questions
Research
Questions
Research
Questions
TrustedStore(Trustworthy Information Store)
L3:
L2:
L1:
P1⇄ P2⇅P1⇄ P3❖P2⇅P1⇄
Figure 2: We group research towards accountable and dependable manufacturing into three layers.L1: Blockchain-inherent challenges that concern the properties of blockchain technology which is
expected to serve as an underlying key component of our envisioned TrustedStore.L2: Scenario-driven challenges that can be grouped into three main research directions that each focus
on a specific research pillar, i.e., along supply chains (P1 Õ), across supply chains (P2 Ö), andsituations with insufficient trust between stakeholders (P3 v).
L3: Socio-economic challenges that have an impact on underlying collaborations and improvements.To offer viable solutions for accountable and dependable manufacturing, research must consider andtackle all layers and their individual research challenges.
infrastructure. To this end, any proposedarchitecture must take the deployment andoperation costs into account, with a special focuson computing overhead for securely keepingdata on-chain. Improvements in efficiencymainly originate from more fundamental linesof research, e.g., advances in authentication,distributed consensus, or secure communication.Yet, a proper integration of these advancesinto a full blockchain-based architecture ismandatory to seize this potential for efficientdata management and to not undermine any
requirements of the overall system. The mainbottleneck of traditional blockchains is theredundant execution of various tasks, suchas verifying digital signatures or maintaininga local state [5]. This redundancy not onlyincreases costs but also creates a potentiallyavoidable environmental impact. Solutions, suchas sidechains or sharding [5], that distribute theworkload without lowering security guaranteeswill help to reduce the operating costs. Whilethese concepts are primarily being researched forpublic settings, the envisioned high-frequency
6 Computer
utilization and large volumes of data call forsimilar developments for consortium blockchains.
Immutability Recording events immutably de-spite the presence of adversaries eager to alterhistory is arguably the blockchain’s key achieve-ment. Thus, storing non-financial, application-specific data on-chain or referencing such datathrough on-chain fingerprints, has become a fre-quent proposition [10]. However, this immutabil-ity has also proved to create further issuesthan only impacting the long-term scalability ofblockchains, e.g., distributing and storing un-wanted blockchain data can cause legal liabil-ity [16]. While the prevalence of known identi-ties within consortium blockchain mitigates suchrisks, different stakeholders may nevertheless bein conflict about the value of recorded data, e.g.,whether data is outdated or when unknown rawdata formats pollute the shared storage. Overall,the quality of recorded information becomes moreimportant as participants should be able to relyon data that is recorded by other parties thatexhibit varying individual levels of trust. Today, alink between a physical (product) property and itsdigital data is missing, which limits the consensusalgorithms’ ability to verify claimed events beforepersisting them on-chain, e.g., sensor readingsfrom inaccessible, remote environments. Correct-ing identified errors is trivially possible by over-writing data in a new transaction, but impliesa more complex transaction processing by allparties. Hence, further research is required toexplore the trade-off between data availabilityand data utility as well as data verifiability andefficient corrections.
Privacy Tightly related to the individual datavalue for different stakeholders involved in theconsortium blockchain is the notion of data pri-vacy, which applies not only to traditional privacy,e.g., storing and trading customer data, but toinformation leakage in general [16]. On the onehand, blockchains may disclose sensitive busi-ness secrets [13], such as capabilities of produc-tion machines or process details, e.g., requiredtemperatures or metal alloys, both directly andindirectly. On the other hand, meta-informationsuch as the frequency of transactions betweentwo collaborators or key performance indica-
tors may be inferred, putting affected parties ata disadvantage against competitors, e.g., duringprice negotiations or when company acquisitionis imminent. A key challenge for sustainableconsortium blockchains will be carefully gaugingthe desired level of point-to-point collaborationsand consequently tackling arising trust barriersthrough both trust and data management.
Scenario-Driven Research Directions (L2)On top of the blockchain-inherent challenges,
further research directions may lead to a fast, ver-satile, accountable, and dependable blockchain-backed manufacturing (cf. Figure 2). Researchinto (i) reliable product information will en-sure the availability of high-quality data along-side all production steps of a supply chain (P1Õ), ranging from tamperproof sensing to se-cure blockchain storage. Based on this reliable,high-quality information more (ii) efficient anddependable collaborations can form in the fu-ture that will increasingly affect dataflows acrosssupply chains (P2 Ö). Ultimately, (iii) dynamicdistributed markets allow for flexible sharing ofdata and advertising services, especially whenstakeholders without any trusted or previous rela-tionships intend to collaborate (P3 v). This way,collaborators can efficiently foster fast, versatile,and dependable business relations.
Reliable Product Information Today, large-scale production and supply chains (P1 Õ) areopaque regarding processes and the origin ofprocessed goods [4]. Consequentially, failure rootcauses and other issues cannot be tracked downefficiently, creating massive administrative over-heads [6], [14], e.g., hampering legal investiga-tions, causing over-dimensioned product recalls,or an inefficient lookup of compatible spare partsfor repairs or assembling bigger workpieces. Sim-ilarly, feeding back information from mid-termor long-term field experience into manufacturingprocesses for improvements is hard [2].
To overcome these limitations, manufacturingneeds a reliably accessible, tamperproof informa-tion store that links clearly identifiable productsto their physical state in a verifiable manner.For example, the transportation of fresh produce,which must uphold a mandated cold chain, re-quires the container’s temperature to be con-
September 2020 7
Feature
tinually monitored such that tricking sensors isinfeasible [8].
First, this process requires measures toachieve a tamperproof gathering of physical-stateinformation. Here, we identify tailored machinelearning mechanisms for anomaly detection aspromising research area. Such a machine learningalgorithm can base on the following data: (i) Us-ing multiple sensors allows for cross-checkinggathered data, e.g., sensors redundantly monitor-ing the container from different vantage pointscan increase tamper resilience as already subtlemonitoring inconsistencies could unveil manipu-lations. (ii) Similarly, different sensor types andmeasuring methods further increase the range forsensing correlation to detect anomalies regard-ing the coherence of real-world physical effects.As sensor nodes cheapen and allow for long-lasting battery-based operation, these solutionsare also becoming increasingly economically vi-able. (iii) Further, high sampling rates also im-prove tamper resilience, as more readings areavailable to identify inconsistencies. Overall, thegathered data provides promising input for amachine learning-based anomaly detection.
Still, storing these large amounts of raw data(i–iii) in globally replicated tamperproof storagessuch as the blockchain remains challenging. In-stead, we envision a combination of mid-termlocal storages maintained by companies and along-term distributed information store. In thisdeployment model, companies store their rawproduction data locally and signal its availabilityon-chain via fingerprints. Further, the blockchainstores (small-sized) insights that result from anal-yses of the locally stored raw data. Likewise,this storage happens in a certified manner, overallcreating a trustworthy information store, whichwe refer to as TrustedStore. To ensure that com-panies fully preserve raw data locally, certifiedservice providers (verifiers) periodically check iflocal stores match with the TrustedStore, so thatmisbehavior can be detected in a timely mannerand appropriately acted upon (legally). As theamount of data renders full-blown checks imprac-ticable from remote locations and on-site checksinvolve high costs, they have to happen onlyrarely. In between, verifiers remotely request datafor randomly selected fingerprints to frequently,yet economically, check for data availability. Al-
ternatively, companies store raw data in globallydistributed certified data stores and prove suchstorage to the TrustedStore. Overall, decouplingthe storage of large amounts of raw data fromderived insights and key properties ensures theimmutability and availability of rich raw datawhile keeping reasonable loads for globally main-tained infrastructures.
Second, a tamperproof digital processing ofgathered data ensures that original sensor read-ings enter the blockchain-backed TrustedStorecorrectly. This way, data can be collected evenfrom untrusted or hostile environments, e.g.,to realize new collaborations without sufficienttrust levels. Tamperproof sensors can providethis form of dependable data gathering and pro-cessing [17]. Such devices combine traditionalsensors, e.g., RFID scanners, or temperature orhumidity sensors [18], with trusted computingmechanisms, such as hardware security modules(HSMs). These security-enhanced sensors areable to immediately hand over data to HSMs forprocessing, thereby minimizing the attack surfacefor tampering. Ultimately, the HSM uploads thesensor readings to the local storage and storestheir fingerprints on the TrustedStore. From thispoint on, the reliably-sensed data is persistedimmutably.
Assuming mechanisms for tamperproof sens-ing and blockchain inclusion, we finally mustclearly link these readings to the respective phys-ical products, e.g., via camera tracking, RFIDtags, imprints, or other markings. Importantly,this identification must also be tamperproof, usingsuitable mechanisms as described before.
In summary, this research will yield a reliablyaccessible, tamperproof TrustedStore for produc-tion data to establish accountability along anysupply chain. Beyond aiding legal investigation,managing product recalls, and optimizing partsutilization, this TrustedStore can further serve asa medium to foster collaborations among well-known and novel companies alike.
Efficient and Dependable CollaborationEstablished business relations with trust in placecan increase their efficiency with a dependableTrustedStore. This claim especially holds fordataflows across supply chains (P2 Ö) that couldimprove the productivity in manufacturing qual-
8 Computer
ity [2]. Additionally, sharing workpiece data,production machine schedules, and states in atimely manner enables close collaborations, ac-cumulating companies into digital factories withproduction efficiencies similar to single, multi-factory companies. Rich information flows allowfor a cross-company allocation of machine timeand flexible handling of process deviations [2],e.g., by automatically reallocating machine ca-pacity in case of delays. Here, the TrustedStoreenables trustworthy tracking methods for work-pieces along the full (multi-factory) supply chain.As a result, problems can easily be tracked, andclearly assigned responsibilities motivate partic-ipants to comply with their obligations. Mostbasically, this information allows for detectinginfringements early on, e.g., misconfiguration ormaintenance backlogs.
Beyond supply chain management, Trusted-Stores simplify the billing of goods or ma-chine usage (Manufacturing-as-a-Service) [2]. Es-pecially with production environments shiftingfrom generic mass production to individual prod-ucts, companies require verifiable and highly au-tomated payment processes to keep administrativeburdens at a reasonable level. Even pay-as-you-go contracts for cost-efficient machine usage inadaptive production are conceivable where cus-tomers pay only for the resources and energyrequired to create the requested (potentially low-quantity) workpieces. Thereby, high degrees ofautomation enable manufacturers to maintain ahigh utilization as multiple customers can sharesingle machines with almost no downtime.
Managing data from mid-term and long-termfield experience on the TrustedStore promisesfurther benefits. In contrast to the previouslydiscussed less sensitive product data, the processdata considered here is more valuable and, thus,must be protected accordingly. Nowadays, infor-mation on product life cycles, required mainte-nance intervals, or production quality variations isexclusively accessible to the manufacturer. Usingthe TrustedStore, such data becomes accessibleto current and prospective machine users alike(cf. Figure 1). Here, the TrustedStore providesevidence of data correctness. Data of individualmachines further facilitates reselling as prior us-age and output quality become assessable.
Research has to answer questions on the
required granularity of sharing data to achievethese envisioned benefits. As business secrets arepotentially at risk when providing informationto external, partially trusted collaborators [2],companies have to make informed decisions whentrading off efficiency and profit for data privacy.
Dynamic Distributed Markets Ultimately,we envision (distributed and transparent)blockchain-based bidding platforms that realizefast, versatile, yet dependable markets for goods,services (e.g., machine rentals), and configurationknowledge, especially fostering collaborationsbetween–previously unknown and potentiallyuntrusted–business partners (P3 v). Today’sbusiness relations typically evolve over longperiods and trust builds up slowly or is enforcedthrough complex contracts. Blockchains canlargely substitute social trust through technicalguarantees and thus foster the establishment ofnew business relations. Furthermore, a distributedTrustedStore allows for efficient automation,e.g., the allocation of machine time, achievinghigh utilization even in adaptive manufacturingprocesses. Consequentially, manufacturers cangenerate profit even from short-time businessrelations for single workpieces, which wouldotherwise be uneconomical and incur high risks.
Customers can search for the best-matchingoffer and benefit from reasonable prices due toincreased market competition. Especially smallermanufacturers can profit from low-barrier mar-ket access to appeal to customers and businesspartners and easily increase (domain) knowledgethrough the TrustedStore.
However, the realization of these distributedmarkets faces a big challenge, i.e., the potentialdisclosure of business secrets. For example, bigcompanies could exploit the TrustedStore’s infor-mation to suppress competitors, e.g., by engagingin well-informed price dumping. Thus, a funda-mental research question is how to match businesspartners based on desired capabilities and quality-guarantees without requiring manufacturers to re-veal too sensitive information up front. Promisingbuilding blocks for such a privacy-preservingcatalog are known from privacy-preserving com-puting. However, they require extensive researchto fit the desired scenario of privacy-preservingbidding platforms for manufacturing.
September 2020 9
Feature
Such mechanisms must realize fair data shar-ing, i.e., participants must not obtain detailedinformation about other participants, especiallycompetitors, without providing said informationthemselves. To this end, mechanisms to assess thevalue of data can provide measures to rate-limitor charge participants with extraordinary usagepatterns.
Socio-Economic Challenges (L3)Beyond the outlined technical measures to
realize accountable and dependable manufactur-ing, we also briefly discuss overarching socio-economic challenges (cf. Figure 2).
Legal Frameworks Legislation currently failsto cover blockchain-based smart contracts andanalyses have to show whether general rulessuffice to enable the envisioned business rela-tions. Especially when considering global supplychains, also different legal frameworks and multi-national agreements must be taken into account.To realize the desired accountability, legal frame-works must further clarify the responsibility forthe accuracy of information in a TrustedStore.An exemplary question is whether manufacturersshould be responsible only for the data they pro-vide or whether they should also be responsiblefor consistency checks on the received data.
In terms of privacy, all systems must complywith local as well as multi-national rules for dataprivacy, such as the GDPR, including the rightto erasure of previously recorded data. Thus, anextensive analysis has to show which data is safeto be stored on-chain, and systems must preventthe inclusion of data that falls under the right to beforgotten or provide mechanisms for data removalwithout undermining the desired goals.
Furthermore, several third-party services thatuse the available data are conceivable, e.g., uti-lizing individual usage data to offer improvedmaintenance for all customers. To this end, legalframeworks have to clarify who owns the dataon the blockchain and who is allowed to processwhich data in which way. Similar questions alsoarise for any derived knowledge.
Access and Transparency Before realizingimmutable TrustedStores, research must work outthe access requirements for different entities and
the corresponding trade-off between verifiabilityand privacy. On the one hand, broad accessto information increases transparency such thatcustomers can obtain information more easily.Research must reveal which information is nec-essary, e.g., to alleviate the required trust fromtoday’s slowly forming business relations viatechnical measures to ease collaboration withoutpre-established trust. Legal entities may furtherdemand access, e.g., to discover cartels.
On the other hand, information stored ona (semi-)public blockchain must not subvertprivacy-legislation. Specifically, granting broadaccess to information may put business se-crets and privacy at risk. Furthermore, reason-able freedom of action for market participantsmust be maintained. For example, adequate mea-sures must prevent customers from exploiting theknowledge of a participant’s low machine utiliza-tion to achieve an uneconomic price. In the end,socio-economic research must develop guidelinesfor blockchain-based platforms that do not onlyoptimize cost but lead to a healthy ecosystem withincentives for high quality, economically healthycompanies, and employee well-being.
FUTURE manufacturing will be driven by ex-citing advances stemming from the combinationof IoT and blockchain technology to implementa dependable and accountable ecosystem. Weidentified relevant future use cases for both supplychain-related and unrelated aspects that shouldsignificantly improve the utilization of manufac-turing data (cf. Figure 1). In particular, researchmust address open challenges on different layers,ranging from system-specific blockchain ques-tions to overarching socio-economic challenges(cf. Figure 2). Regardless, we believe that mosteffort must be invested in scenario-driven tasks toenable trustworthy information stores, i.e., Trust-edStores, in competitive, business-driven, and po-tentially distrustful industry environments. Fortu-nately, smaller advances are already achievable inincrements, and as such first changes should berealizable in the near future.
ACKNOWLEDGMENTFunded by the Deutsche Forschungsgemein-
schaft (DFG, German Research Foundation) un-der Germany’s Excellence Strategy – EXC-2023Internet of Production – 390621612.
10 Computer
REFERENCES
1. A. Kusiak, “Smart manufacturing must embrace big
data,” Nature, vol. 544, no. 7648, pp. 23–25, 2017.
2. J. Pennekamp, M. Henze, S. Schmidt, P. Niemietz,
M. Fey, D. Trauth, T. Bergs, C. Brecher, and K. Wehrle,
“Dataflow Challenges in an Internet of Production: A
Security & Privacy Perspective,” in Proceedings of the
ACM Workshop on Cyber-Physical Systems Security &
Privacy (CPS-SPC ’19). ACM, 2019, pp. 27–38.
3. S. Malik, S. S. Kanhere, and R. Jurdak, “ProductChain:
Scalable Blockchain Framework to Support Provenance
in Supply Chains,” in 2018 IEEE 17th International
Symposium on Network Computing and Applications
(NCA). IEEE, 2018, pp. 20:1–20:10.
4. L. Gleim, J. Pennekamp, M. Liebenberg, M. Buchs-
baum, P. Niemietz, S. Knape, A. Epple, S. Storms,
D. Trauth, T. Bergs, C. Brecher, S. Decker, G. Lake-
meyer, and K. Wehrle, “FactDAG: Formalizing Data In-
teroperability in an Internet of Production,” IEEE Internet
of Things Journal, vol. 7, no. 4, pp. 3243–3253, 2020.
5. R. Matzutt, B. Kalde, J. Pennekamp, D. Arthur,
M. Henze, T. Bergs, and K. Wehrle, “How to Securely
Prune Bitcoin’s Blockchain,” in 2020 IFIP Networking
Conference (Networking). IEEE, 2020, pp. 298–306.
6. S. Wang, D. Li, Y. Zhang, and J. Chen, “Smart Contract-
Based Product Traceability System in the Supply Chain
Scenario,” IEEE Access, vol. 7, pp. 115 122–115 133,
2019.
7. M. Montecchi, K. Plangger, and M. Etter, “It’s real,
trust me! Establishing supply chain provenance using
blockchain,” Business Horizons, vol. 62, no. 3, pp. 283–
293, 2019.
8. S. Malik, V. Dedeoglu, S. S. Kanhere, and R. Jurdak,
“TrustChain: Trust Management in Blockchain and IoT
supported Supply Chains,” in 2019 International Con-
ference on Blockchain (Blockchain). IEEE, 2019, pp.
184–193.
9. K. Korpela, J. Hallikas, and T. Dahlberg, “Digital Supply
Chain Transformation toward Blockchain Integration,” in
Proceedings of the 50th Hawaii International Confer-
ence on System Sciences (HICSS). AIS, 2017, pp.
4182–4191.
10. M. Bartoletti and L. Pompianu, “An analysis of Bitcoin
OP RETURN metadata,” in International Conference
on Financial Cryptography and Data Security (FC).
Springer, 2017, pp. 218–230.
11. E. Mengelkamp, J. Garttner, K. Rock, S. Kessler,
L. Orsini, and C. Weinhardt, “Designing microgrid en-
ergy markets: A case study: The Brooklyn Microgrid,”
Applied Energy, vol. 210, pp. 870–880, 2018.
12. Z. Yang, K. Yang, L. Lei, K. Zheng, and V. C. Leung,
“Blockchain-Based Decentralized Trust Management in
Vehicular Networks,” IEEE Internet of Things Journal,
vol. 6, no. 2, pp. 1495–1505, 2019.
13. Q. Lin, H. Wang, X. Pei, and J. Wang, “Food Safety
Traceability System Based on Blockchain and EPCIS,”
IEEE Access, vol. 7, pp. 20 698–20 707, 2019.
14. J. Pennekamp, L. Bader, R. Matzutt, P. Niemietz,
D. Trauth, M. Henze, T. Bergs, and K. Wehrle, “Pri-
vate Multi-Hop Accountability for Supply Chains,” in
2020 IEEE International Conference on Communica-
tions Workshops (ICC Workshops). IEEE, 2020, pp.
1–7.
15. C. Cachin and M. Vukolic, “Blockchain Consensus Pro-
tocols in the Wild,” in 31st International Symposium on
Distributed Computing (DISC 2017). Schloss Dagstuhl,
2017, pp. 1:1–1:16.
16. R. Matzutt, J. Hiller, M. Henze, J. H. Ziegeldorf,
D. Mullmann, O. Hohlfeld, and K. Wehrle, “A Quantita-
tive Analysis of the Impact of Arbitrary Blockchain Con-
tent on Bitcoin,” in International Conference on Financial
Cryptography and Data Security (FC). Springer, 2018,
pp. 420–438.
17. J. Pennekamp, F. Alder, R. Matzutt, J. T. Muhlberg,
F. Piessens, and K. Wehrle, “Secure End-to-End Sens-
ing in Supply Chains,” in 2020 IEEE Conference on
Communications and Network Security (CNS). IEEE,
2020, pp. 1–6.
18. F. Tian, “An Agri-food Supply Chain Traceability System
for China based on RFID & Blockchain Technology,” in
2016 13th International Conference on Service Sys-
tems and Service Management (ICSSSM), 2016, pp.
8:1–8:6.
Jan Pennekamp received his B.Sc. degree andM.Sc. degree in Computer Science from RWTHAachen University with honors. He is a researcherat the Chair of Communication and DistributedSystems (COMSYS) at RWTH Aachen University,Germany. His research focuses on security & pri-vacy aspects in the Industrial Internet of Things(IIoT). He is IEEE Student Member. Contact him [email protected].
Roman Matzutt received his B.Sc. degree andM.Sc. degree in Computer Science from RWTHAachen University. He is a researcher at the Chairof Communication and Distributed Systems (COM-SYS) at RWTH Aachen University, Germany. His re-search focuses on blockchain and its privacy impli-
Salil S. Kanhere received his M.S. degree and Ph.D.degree from Drexel University in Philadelphia. He isa Professor of Computer Science and Engineeringat UNSW Sydney, Australia. His research interestsinclude Internet of Things, blockchain, pervasive com-puting, cybersecurity and applied machine learning.He is a Senior Member of the IEEE and ACM and anHumboldt Research Fellow. He serves as the Editorin Chief of the Ad Hoc Networks journal. Contact himat [email protected].
Jens Hiller received his B.Sc. degree and M.Sc.degree in Computer Science from RWTH AachenUniversity. He is a researcher at the Chair of Com-munication and Distributed Systems (COMSYS) atRWTH Aachen University, Germany. His researchfocuses on efficient secure communication in the In-ternet of Things. Contact him at [email protected].
Klaus Wehrle received his Diploma (equiv. M.Sc.)and PhD degree from University of Karlsruhe (nowKIT), both with honors. He is full professor at theChair of Communication and Distributed Systems(COMSYS) at RWTH Aachen University, Germany.His research interests include network protocol en-gineering, methods for network analysis, and reliablecommunication. He is a Member of IEEE and ACM.Contact him at [email protected].
FURTHER READINGWe provide references to further reading ma-
terial related to this article for an overview intorelated work and today’s relevant research chal-lenges. In particular, our selected literature pro-vides additional insights into challenges (1) andapplication areas (2–4) of blockchain technologyas well as supply chain-specific research (5–6).Finally, we include literature on the envisionedInternet of Production (7) and associated chal-lenges when processing big data (8).
1) Blockchain challenges:Z. Zheng, S. Xie, H.-N. Dai, H. Wang andX. Chen, “Blockchain Challenges and Op-portunities: A Survey,” Inderscience, 2018,International Journal of Web and Grid Ser-vices, vol. 14, no. 4, p. 352–375.
2) Financial blockchain applications:Y. Guo and C. Liang, “Blockchain applica-
tion and outlook in the banking industry,”Springer, 2016, Financial Innovation, vol. 2,no. 1, p. 24:1–24:12.
3) Survey of blockchain-based applications:F. Casino, T. K. Dasaklis, and C. Pat-sakis, “A systematic literature review ofblockchain-based applications: Current sta-tus, classification and open issues,” Elsevier,2019, Telematics and Informatics, vol. 36,pp. 55–81.
4) Determining the suitability of blockchain:K. Wust and A. Gervais, “Do you need aBlockchain?” in Crypto Valley Conferenceon Blockchain Technology (CVCBT). IEEE,2018, pp. 45–54.
5) Blockchain-supported use cases in thecontext of supply chains:P. Gonczol, P. Katsikouli, L. Herskind, andN. Dragoni, “Blockchain Implementationsand Use Cases for Supply Chains-A Sur-vey,” IEEE, 2020, IEEE Access, vol. 8, pp.11 856–11 871.
6) Research propositions for supply chains:A. Rejeb, J. G. Keogh, and H. Treiblmaier,“Leveraging the Internet of Things andBlockchain Technology in Supply ChainManagement,” MDPI, 2019, Future Internet,vol. 11, no. 7, pp. 1–22.
7) Internet of Production:J. Pennekamp, R. Glebke, M. Henze,T. Meisen, C. Quix, R. Hai, L. Gleim,P. Niemietz, M. Rudack, S. Knape, A. Ep-ple, D. Trauth, U. Vroomen, T. Bergs,C. Brecher, A. Buhrig-Polaczek, M. Jarke,and K. Wehrle, “Towards an InfrastructureEnabling the Internet of Production,” in 2019IEEE International Conference on IndustrialCyber Physical Systems (ICPS). IEEE, 2019,pp. 31–37.
8) Challenges in Big Data:A. Oussous, F.-Z. Benjelloun, A. A. Lah-cen, and S. Belfkih, “Big Data technologies:A survey,” Elsevier, 2018, Journal of KingSaud University-Computer and InformationSciences, vol. 30, no. 4, pp. 431–448.
