As we approach the end of this decade, I thought it would be worth having an honest and impartial reflection on how enterprise Blockchain technology usage has evolved over the past five years, since the very first release of Ethereum and smart contracts, and examine where we’ve landed as an industry.
It’s fair to say that it has been a very exciting and turbulent time, and a great privilege to be working in the space as this transformative new technology finds its footing.
There are many different potential applications of Blockchain technology, and I’m going to examine the space from an enterprise and business technology perspective, as this is the market in which we operate.
Enterprise, private, consortium blockchains
The enterprise Blockchain space is broadly divided into three major camps:
- The purists who believe in the original Ethereum vision of a global public Blockchain, and promote private Ethereum implementations including Quorum in the enterprise space with a view that one day these two worlds will connect.
- The pure enterprise players who wrote off Ethereum technology as a cryptocurrency play unsuitable for enterprise, and see Hyperledger Fabric as the only viable option for enterprise.
- The R3 Corda camp, evolved by banks through their initial engagement, and subsequently through many other players who view Corda as a great way to be able to connect their applications to the banks and each other.
I’d like to explore their fundamental architectures, and the major applications that have been built in the enterprise space.
There are of course thousands of different PoC’s and Blockchain applications, but it appears that the most significant investments have been in trade and trade finance (see the recent WTO report). This is where we see the largest, and best funded group of consortia (TradeLens on Fabric, MarcoPolo on Corda, WeTrade on Fabric, VAKT on Quorum, KOMGO on Quorum etc.). The reason for this is that this is relatively low hanging fruit. The industry was largely decentralised, paper based, slow, and high value, and Blockchain technology provided the perfect backdrop for partners and competitors to come together, digitise their world and make it more efficient.
But how is Blockchain actually being used in these platforms? A clue to the answer can be found in the architecture of the technology platforms they are using.
These solutions are first and foremost about business process. They are about document digitisation and process automation. They are about being in sync and reducing the potential for fraud with counterparts. They are about cross-party enterprise integration. And they are very good at solving these problems.
It took me a while to realise this, but these solutions cater for relatively short-lived multi-party transactions. And if you look at how Hyperledger Fabric has evolved (from its original incarnations which looked very much like Ethereum), and how Corda and even Quorum are architected, they are very good at servicing private transactions within small groups.
This is fantastic, and giant step forward for enterprise integration.
I sometimes hear people criticise enterprise blockchain technology claiming: “it’s just a distributed database”. That’s right! Multi-party distributed databases are hard, and we they didn’t actually have mainstream adoption until blockchain.
However, what these are not is asset based platforms. Nor are they optimised for long running provenance of assets across different groups of transacting parties.
The original public Blockchain implementations like Ethereum were designed for trading assets. The ERC-20 token is an example of how this is engineered in practice. However, in adapting blockchains for enterprise, and in particular for data privacy and scalability in business process focused use cases we’ve lost some of that original promise and capability, and it’s very important to appreciate that.
To be fair, Fabric, Quorum and Corda all have asset tracking capabilities, but these are an afterthought in technical architecture terms, where the primary considerations when they were first designed were data isolation, data privacy and scalability.
These platforms can arguably be used with no isolation nor channels, in which case they could track assets and provide provenance, but this undermines their primary privacy model. The separation of channels in Fabric, private transaction managers in Quorum, and individuals ledgers in Corda means that each private transaction group is effectively a different Blockchain, and any assets that would need to move across these privacy boundaries would need to be “carried” across with all of their proofs of history and provenance. While this is technically feasible, it is not particularly efficient, especially when compared to blockchains that are optimised for asset ownership and transfer.
This difference in focus and capability around business processes vs assets also explains the differing approaches to data on chain.
Enterprise applications built using Hyperledger Fabric and Corda typically use the blockchain as a shared database. These blockchains store data that needs to be shared between parties and processed by smart contracts.
Conversely, asset-focused blockchains tend to minimise the storage of data on chain, and only store minimal information to identity the asset, the owner, and execute the smart contract rules that govern asset movement.
Similarly, the difference in focus and capability around business processes vs assets also explains the differing security and decentralisation requirements.
Blockchain networks acting as asset ownership registries require a high level of security through extensive decentralisation to minimise the possibility of collusion, fraud and theft. The monetary value of compromised assets could be very high.
Conversely, business process focused blockchains, while still security conscious tend to have less at stake, when compared to asset registries, and it is therefore more acceptable for them to be secured by much smaller groups of validators in channels, or only by the parties to a particular transaction.
I am generalising. There’s more to security than decentralisation, and there are of course large highly regulated entities that act as custodians of assets today and would be perfectly comfortable in a small group perhaps with a third party acting as a notary.
I raise all of this because of perceptions and expectations. Many of us originally understood blockchain and its potential capabilities and benefits through examples of how public blockchains such as Bitcoin and Ethereum work, and then assumed that these overall mechanics and benefits also apply to the enterprise blockchain platforms only in a way that is compatible with enterprise IT requirements. That’s not really the case right now. The differences may seem subtle, but the implications may be quite substantial, depending on the use case and ambitions.
It is worth stepping back, understanding and separating what is possible with Blockchain in general; and what has been designed, architected, implemented and optimised in specific enterprise blockchain platforms, as this is not the same thing.
It’s really a question of where you seek value from blockchain technology. We describe three areas:
- Business process
The enterprise blockchain projects and platforms mentioned above are making great strides in 1: business process. This is the safest place for most enterprises to operate right now, and the major enterprise blockchain platforms are tailored for this and optimised accordingly.
We’re constantly being challenged on the value of enterprise blockchain initiatives, and when viewed in this light we can see clear value in automating cross-organisation business process workflows on a trusted, shared platform.
The second area, where we potentially see more value is asset registration, ownership, and exchange, be that for digital or physical assets. In order to really deliver value the asset record secured on the blockchain should have legal recourse, and there is still much work to do in this regard.
The third area, of money, is where the technology began, but is likely to be the last in terms of enterprise adoption, due to legal, regulatory and a swathe of other obstacles. There are of course a number of central banks, commercial banks, and of course Libra working on this.
Many supply chain use cases require products and components to be treated as “assets” tracked and transferred across different private groups of transacting parties. Most of the automated business processes discussed earlier ultimately result in a transfer of funds; and often in a transfer of assets. In these cases there is an expectation that the enterprise Blockchain platforms will enable digital money and tokenised assets to be integrated with business process through smart contracts.
If these enterprise applications are extended to trade assets and move money, the enterprise platforms will be required to provide an architecture that is optimised for both business processes as well as tokenisation in a private and efficient manner.
In the meantime we’re seeing a fair amount of “platform gymnastics”, where platforms are stretched to be used in a way that they weren’t quite designed, or multiple platforms are used in a single solution, because one blockchain platforms does one thing well (e.g. business process), and another does something else well (e.g. tokenization). This is obviously technically inefficient and cumbersome.
Zero knowledge proofs
Recent advances in cryptography, including zero knowledge proofs open up new possibilities for combining enterprise grade privacy, without the need to isolate, and enterprise platforms are beginning to experiment (see Hyperledger Fabric ZKAT, Quorum ZSL, and even our very own K0).
This may see us move away from private channels and transaction groups, to broader consensus networks that are still able to maintain absolute privacy.
Blockchains that never grow
As well as rewriting the rule book on blockchain data privacy, zero knowledge proofs are being used to improve scalability with blockchains that never grow (see Coda). Along similar lines, Starkware is “rolling up” transactions outside a blockchain that can then be verified very efficiently on chain.
All of the above has led Gartner to predict:
90% of Current Enterprise Blockchain Platform Implementations Will Require Replacement by 2021
While statement is a little extreme, it does reflect the pace of change and optimisation occurring in the underlying technology.
Enterprises and consortia building blockchain solutions must constantly reassess and future proof them in order to protect their investments.
Public smart contract blockchains
Looking across the space, and considering developer and business adoption, we still see one very dominant public smart contract platform: Ethereum. The platform is largely unchanged since its launch. The number of live public deployed business applications is still tiny, but the level of research projects, development activity, PoC’s and private deployments is tremendous, and it is still the de facto public Blockchain smart contact platform that most developers still refer to.
There are now literally hundreds of competitors to Ethereum, including some very interesting platforms such as Hedera and Algorand, all improving some aspect of the technology; whether its scalability through a smaller group of validators, or roll-ups, or privacy using zero knowledge proofs, or a more secure and efficient consensus mechanism, but most have yet to seriously challenge Ethereum for a share of the developer mindset.
There’s something about the purity of the Ethereum model, and major decentralisation thanks to proof of work; the true distribution through fully replicated nodes; and the sheer potential of the Turing complete smart contracts, and the history of projects, frameworks and standards (see ERC-20 and it’s variants) that has enabled it to retain its dominant position.
This is I why think a major enterprise player like EY has chosen to pursue a strategy based on the public Ethereum Blockchain.
Ethereum itself has a roadmap of major transformations through the controversial 2.0 to the already talked about 3.0. Numerous teams are both collaborating and at the same time competing on its development. Time will tell whether these will deliver in way that enables the platform to retain its dominance against increasing competition.
As part of a reality check, we’ve looked not only at where blockchain technology is utilised, but through a more critical lens: how and why.
It’s not just about building applications on blockchain. It’s about knowing how to tap into real value in the most efficient way, leveraging a deep and impartial understanding of the technology.
It is worth reflecting on the fact that we are still in the very early days of a technology evolution, and a comprehensive understanding should encompass not only blockchain but also broader emerging areas of advanced cryptography.