Blockchain in Smart Manufacturing: the Where and the When

Blockchain technology was quite a hype a few years ago. Now things have become a bit quieter – and other technology hypes are around – but this doesn’t mean that blockchain as a technology has become less relevant. It rather means that it has reached a higher level of maturity in the world of digital business enablement. In terms of Gartner’s well-known hype cycle of technologies, blockchain technology is moving from the ‘peak of inflated expectations’ towards the ‘plateau of productivity’. Blockchain technology is interesting for many application domains where business data and trust management meet. Smart manufacturing is one of these domains.

In this short article, we explain the use of blockchain in smart manufacturing from two points of view. Firstly, we outline in which areas of a smart manufacturing scenario blockchain can possibly be used to achieve which goals. Secondly, we explain how you decide whether the use of blockchain is actually a viable idea in an identified scenario. But before we dive into this, let’s briefly agree on what blockchain actually is, such that we start with a common basis – there are some misunderstandings.

What is blockchain?

During its hype period, blockchain has been called a lot of things – about everything from a ‘strategy management paradigm’ to an ‘IT infrastructure for all purposes’. Much of this is exaggerated or simply wrong (to avoid terms like ‘bullshit’ in a well-mannered article like this). Blockchain essentially is a digital technology to realize a trusted distributed ledger. A ledger in this context is an administration of facts – or if you prefer, a log of transaction descriptors. It is distributed because it is kept (or managed) by multiple parties – typically all those involved in a business scenario. It is trusted because it is virtually impossible to tamper with the contents of the ledger. The special feature of blockchain is that you don’t need a trusted third party (like a bank or a notary) to create the trust – the technology enforces it. We’ll not go into the details of this technology here – those interested can consult sources like our recent book on the use of blockchain in business [1]. To summarize things very briefly: blockchain technology allows partners in a business ecosystem to agree on events that have happened between them without involving an expensive trusted third party. This can be the basis for several applications of blockchain in the manufacturing domain.

Where can I use blockchain in manufacturing?

As implied by the above short explanation, blockchain can be used in manufacturing scenarios in which data is shared between different business partners and trust in the correctness of this data is essential. We can classify the various uses of blockchain by considering processes in a manufacturing value chain from the very physical to the more digital in five steps. Let's take the manufacturing of electric trucks as an arbitrary example.

Supporting Goods flow. The most physical process between partners in typical manufacturing value chains is the transport of materials and parts to a manufacturer and the transport of completed products (and spare parts) to the customers of the manufacturer. In traditional scenarios, there is a lot of paperwork going with the various transport channels to record what has been delivered from whom to whom at which time, at which place, and under which circumstances. A lot of documents need to be signed or stamped to ensure the required level of trust between partners when goods change ownership. When an electric truck is delivered to a transport company, this company will need to sign for receipt of the truck as well as for the state it was delivered in. This administration can be moved to a blockchain platform to get rid of all the paperwork. Blockchain allows all participating parties to enter transactions and to check entries made by others, creating transparency between collaborating parties. The standard blockchain mechanisms guarantee that entries made cannot be tampered with. Combining blockchain technology in this context with Internet-of-Things (IoT) technology provides the means to automatically enter transaction data into the blockchain – for example by using bar code, QR code or RFID readers along a supply chain. This increases process efficiency and decreases risks of errors and malicious tampering with supply chain data. In scenarios where product genuineness is important (like spare parts for trucks), it can help with preventing counterfeiting by automatically keeping transparent, auditable traces of all goods flows.

Supporting Product servicing. When we move to a slightly less physical scenario in a manufacturing value chain, we can look at the after-sales service provisioning processes, like keeping malfunctioning products in good working order (possibly as part of a warranty contract) and upgrading products (for example by installing new firmware). Here, it is essential to keep an administration of the life cycle of a product, for example to assess whether warranty still applies or to be sure which spare parts to use. This is traditionally achieved by keeping a physical log of life cycle events of a product – or a digital log by one party. The problem can arise, however, that partners do not agree on the data in the log. Here, blockchain can help by registering all relevant life cycle events as transactions in the ledger, such that the status of a product is always transparent to all involved parties. In our electric truck example, the truck may register automatically how many miles it has been driven every day or week, such that maintenance conditions depending on mileage are always clear. In this example, IoT technology comes into play in the form of sensors in the truck that provide the basic data for blockchain entries. Like in the goods flow case, IoT technology can also be used to register the use of spare parts in product servicing (and check their genuineness).

Enabling pay-per-use. In various markets, we see a development from pay-per-products models to pay-per-use models (sometimes via leasing models, which can be seen as pay-per-right-to-use). Returning to our electric truck scenario, a transport company may pay the manufacturer of trucks that they use not per truck delivered, but per ton-mile that a truck has been actually used. In the aircraft industry, for example, this is becoming a more commonly accepted model. When the business relation between a goods producer and a goods user is based on the actual use of the goods, it becomes obviously essential to agree on the amount of use. This pertains both to measuring the use in a jointly agreed way during business operation (possibly using IoT technology) and to storing usage measurements in a jointly agreed, transparent and trusted way, such that these measurements can be the basis for billing and paying. As you may have concluded by now yourself, blockchain can be a perfect technology to implement such a usage measurement storage, thereby enabling the pay-per-use business model also in markets where trust is not ‘fully pre-installed’.

Enabling pay-per-effect. When we take the pay-per-use model one step further, we arrive at the pay-per-effect model. In this case, the user of a good (like the trucks in our example) do not pay for the use if this use does not contribute to their business performance. If trucks do transport enough goods over a sufficient number of miles (so fulfill the pay-per-use criteria) but break down frequently such that the goods often arrive too late at their destination, they do not fulfill the pay-per-effect criteria.? To implement this as a pay-per-effect model, we have to solve similar issues as in the pay-per-use model. Measuring business effects is often, however, harder than measuring ‘plain’ usage and may be related to multiple performance dimensions. Where dimensions are of a ‘physical’ kind (like speed, weight or temperature), IoT-based sensor technology can be of help to make measurements. Where dimensions are non-physical (like price or customer satisfaction), interfaces may be used to digital systems that contain the data on these dimensions. For business-to-consumer scenarios, measuring customer satisfaction may even involve ‘scraping’ social media to assess the opinion of customers of your products.

Enabling pay-per-outcome. The observations with the pay-per-effect model become even more true when we move another (and final) step further to pay-per-outcome. In a basic pay-per-effect model, the payment is typically based on a characteristic that is directly related to the product (like the frequency of breakdowns of a truck in our example). In a pay-per-outcome model, the effect is defined in terms of the business model (and hence the market) of the goods user: it is determined by business outcomes of the customer [2]. This imposes even higher requirements on the way to measure and store performance details: we are now talking about data in the market of the customer (downstream the value chain), not in the market of the producer.

Now we have seen how blockchain technology can be used in making the administration of a manufacturing supply chain transparent and trusted. We have done this in a spectrum of business models ranging from selling pure products to selling customer outcomes – with increasing possibilities for creating customer intimacy along the spectrum, as shown in the above figure. The observation that you can use a technology like blockchain, however, does not automatically imply that you should use the technology. This depends on several factors that we discuss next.

When should I use blockchain?

If a manufacturing scenario is in one of the five classes discussed above, blockchain technology may be a good basis for data management in the corresponding value chain. Whether it actually is a good basis, depends on seven criteria in several categories (based on the models in [1]). If all criteria are met, there is a final decision to be taken.

Business ecosystem. Criterion #1 is about the overall structure of the business ecosystem: does it consist of multiple autonomous business parties? If all involved parties belong to the same business organization, there should be no trust issue that blockchain technology can solve. If most of the involved parties are individual consumers, using blockchain is usually not applicable (it is a B2B technology). Criterion #2 checks whether trust is not already incorporated in another way in the ecosystem: is a trusted authority lacking in the ecosystem? If there is a trust-providing party for data transfer between partners, using blockchain may be overkill. Criterion #3 checks whether there is indeed decentralized operation in the ecosystem: are multiple partners processing data? If one partner processes all data on behalf of the other partners, using blockchain doesn’t make much sense.

Kind of data. Blockchain mechanisms are based on immutable data: data entered into the chain can never be changed nor removed by any party. Therefore, the question whether immutable data is beneficial in the ecosystem forms Criterion #4. If the answer is no, using blockchain becomes questionable. Blockchain mechanisms are computationally intensive and therefore involve certain costs, depending on the amount of data to be stored in the chain. Therefore, Criterion #5 is: is the transaction data small enough, or can it be made small enough? For example, storing complete product specification files in a blockchain may not be a good idea. But possibly the data can be made small enough in this case by storing the correct identifiers only in the chain and storing the rest off-chain.

Kind of data processing. As blockchain mechanisms are computationally intensive, they take some time. Therefore, Criterion #6 asks: can some processing delays be tolerated? In practice, this means that strictly hard-real-time processes may be less suitable for blockchain. These are not so common in manufacturing value chains, however. Blockchain mechanisms are usually not free, so Criterion #7 is: can the value chain transaction costs be increased by the blockchain transaction costs from a business point of view? In practice, this means that blockchain may not be a good idea to manage very small transactions in the value chain. There are technical solutions that can alleviate issues with Criteria #6 and #7 to some extent, for example blockchain platforms that have specifically been designed for an IoT context.

Level of transparency and openness. There are two classes of blockchain solutions: public and private solutions. ?Public blockchains are in principle open to participation and transparent with respect to all data stored. Private blockchains have controlled participation and include mechanisms for control access to parts of the data stored. Depending on the precise application, one class is typically preferable.

The final word

Blockchain is a digital technology that may be relevant to manufacturing scenarios. How it may be relevant and under which circumstances it can be relevant is briefly described above (and more extensively in [1]). As with so many things on the interface of business and digital technology, using blockchain properly is the result of a careful analysis and design process that involves both business and technology experts.

Thanks go to @Marco Comuzzi of @UNIST for proofreading this article and providing feedback.

[1] Comuzzi, M., Grefen, P., Meroni, G. Blockchain for Business: IT Principles into Practice. Routledge Publishers, 2023. ISBN 978-1-03-234246-7. https://www.routledge.com/Blockchain-for-Business-IT-Principles-into-Practice/Comuzzi-Grefen-Meroni/p/book/9781032342467.

[2] Grefen, P.; Vanderfeesten, I.; Wilbik, A.; Comuzzi, M.; Ludwig, H.; Serral, E.; Kuitems, F.; Blanken, M.; Pietrasik, M. Towards Customer Outcome Management in Smart Manufacturing. Machines 2023, 11, 636. https://doi.org/10.3390/machines11060636 (available in open access, free of charge).