Why the Internet of Things needs blockchain technology
Republished by kind permission of CoinTelegraph, where this article first appeared
“#InternetOfThings is when your toaster mines bitcoins to pay off its gambling debts to the fridge.”
When this was tweeted by Andrew Miller in 2015, it rapidly became the go-to meme for anyone slightly cynical about the genuine use cases for blockchain within the Internet of Things.
The idea of a self-actualised toaster with its own digital identity (probably provided by 21 Inc) became as much part of blockchain popular culture as BitFury’s Bitcoin-mining lightbulb that never was or Slock.it’s self-hiring bicycle.
If we are most likely to understand the Internet of Things by describing how it relates to objects we use every day, such as toaster or bicycles, we miss the big picture: a futuristic global network of connected devices, transforming industrial and business processes in a way that we cannot yet comprehend.
The intersection between IoT and blockchain is a fascinating area. Gartner have predicted there will be 20.4 billion connected devices by 2020; and smaller, more efficient sensors and microprocessors offer the potential for use cases of which we can barely dream today. Decentralised architectures mitigate against single points of failure, while providing standard protocols for devices to discover each other and communicate.
For example, imagine a smart device connected to an oil pipeline. With access to historical and current data, smart contracts could be executed which could adjust the flow of oil according to up-to-the-minute global demand, or reroute the oil to areas where adverse weather may be about to stimulate a sudden spike in usage.
In order to do this, the pipeline must be able to consume data from many different sources — and to trust this data. Oil companies would rapidly go out of business if the supply was shut off at the wrong time, or diverted to areas of high demand. This interoperability and trust are key to many IoT interactions.
Take, for example, traffic management systems of the future where autonomous vehicles manufactured by different automotive companies may need to share not only their position but also telemetric data such as energy efficiency or the number of passengers in the car with other vehicles, or with third-party monitoring systems.
Blockchain — or in looser terms, distributed ledger tech — has been touted as the solution to these challenges. It is important to make this distinction between true blockchains and other types of public or private distributed ledgers, as — apart from some services that are being developed using the Bitcoin, Ethereum or other public blockchains — many of these new industrial networks or supply-chain products use technologies whose architecture has only loose similarities with blockchain as we know it.
Four of the most interesting areas where blockchain technology could be useful for the Internet of Things are: supply-chain management; devices selling capacity (or other services) to other devices; devices selling services to humans in the shared economy; and devices sharing data with each other, or with third parties.
Let’s look at each in turn.
Supply-chain tracking and provenance
When people talk or write about the Internet of Things, we tend to talk about devices that are permanently connected. Yet the term IoT theoretically covers any product that can be assigned a digital identity and can be tracked and verified, which could be anything from a sneaker to a tuna (real-life proofs-of-concept by Chronicled and Provenance, respectively).
The two examples above provide proof that an object is what it is claimed to be: an authentic, non-copyright fashion item, or an ethically produced food source, while Chronicled’s Temperature Logger product is designed to certify, for example, that a shipment of meat has been kept below a controlled temperature for the duration of its journey.
While the contract that ensures the provenance of a product may be tamper-proof when written to the blockchain, the same may not be entirely true of the mechanism that is used to link the contract to the real-world object. NFC or BLE chips can be destroyed or removed, and there may need to be some kind of dependency on a centralised authority to verify authenticity at the point of origin, thus removing some of the trust advantages of blockchains.
However interesting this data may be for consumers, it is when these supply chain solutions are scaled to industral levels that they become truly interesting and time-saving.
Filament are a startup working in this space, and have been funded to the tune of $15M. Their approach to industrial networks leverages the power of the Bitcoin blockchain without the costly overhead of registering every single machine-to-machine transaction. As CEO Allison Clift-Jennings: “Some call this an on-device, off-chain smart contract platform. I like that definition — though it’s important to clarify that these contracts do not run in a blockchain — they merely verify inputs and outputs via a blockchain.”
It’s not just startups who are innovating in this area.
Dirk Slama, chief alliance officer at Bosch Software Innovations, said recently: “We are seeing tremendous potential for the application of blockchain in industrial use cases. Being able to create a tamperproof history of how products are manufactured, moved and maintained in complex value networks with many stakeholders is a critical capability. This must be supported by a shared blockchain infrastructure and an integrated Internet of Things protocol.
It’s worth noting that Bosch presented, earlier this year, a blockchain-verified solution for ensuring the distances recorded on vehicle odometers, to prevent insurance and vehicle fraud.
Similarly, when shipping giant Maersk teamed up with IBM for their Cross-Border Supply Chain experiment, it was a statement of intent that even companies on this scale are prepared to innovate if it suits them.
It has been estimated that a large ship of containers can generate as much as 1kg of written paperwork which needs to be signed off manually — a tedious overhead that is ripe for automation.
IBM have also been working with Finnish startup Kouvola Innovation in this area.
Automated ships, their cargo passing seamlessly from country to country, untouched by human hands and recorded in cryptographically secured ledgers as their sensors announce their presence: few would bet against this as the model for global trade a decade from now.
Devices selling or buying capacity from each other
While toasters and fridges mining Bitcoin was never likely to be a productive use case, there are some far more interesting prospects for blockchain-connected devices.
Brooklyn’s Transactive Grid project, powered by LO3 Energy and Siemens, distributes solar power between neighbours. It makes sense for houses with solar panels on the sunny side of the street to be able to sell their excess energy to their less sunny neighbours.
With traditional energy networks and banking systems, the friction and sheer amount of work involved in these transactions mean that no one but the most committed would attempt such a redistribution. Standalone devices for measuring output from each solar panel, combined with a separate invoicing system and the work required for neighbours to chase traditional bank-to-bank payments from each other would make such transactions prohibitively inefficient.
Hence devolving this responsibility to smart contracts makes sense.
Similarly, as any householder knows — especially in the UK — one of the most tedious jobs is using sites like Uswitch to figure out whether you could be paying less for gas, television, broadband, etc. The firms all compete with attractive introductory deals which become less appealing as time goes on. Accenture’s innovation team have developed a smart plug which could find and negotiate the cheapest rate at any given moment, thus obviating the need to sign up to a one-time only deal with an energy supplier, and reducing costs for households.
This raises the possibility of being able to top up our possessions with a certain amount of cash and pass on to them the responsibility of spending it wisely.
Generally, there is a trade-off between the level of immutability (difficulty of tampering) and the speed of the network. The Tangle architecture proposed by IOTA is a blockless network that is nevertheless peer-to-peer, but without the energy costs of a traditional Proof of Work model. The Tangle nodes validate other transactions in return for their own inclusion in the network, but do not have to mine or otherwise earn rewards (all IOTA tokens are pre-mined).
Carsten Stöcker, a senior manager at German energy giant Innogy SE and also an IOTA Foundation member, describes the Directed Acyclic Graph (DAG) structure thus: “When a device makes a transaction and broadcasts the transaction to the network, by protocol or ‘by default’ it has to validate two previous transactions. So, it’s a real peer-to-peer network by machines for machines.”
Prediction: it is only a matter of time before we trust the smart devices in our households to run themselves, to find the best deals for us, and to interact with our neighbours’ households to buy their surplus capacity and services.
The sharing economy
If the average car is used only three per cent of the time, it’s unsurprising that vehicle sharing has been at the forefront of the social change that has seen millennials deserting car ownership. But the trend that could change the face of our cities by turning unsightly car parks into green shared spaces has largely been spearheaded by centralised, private entities. Network effects and investment in infrastructure means that we are more likely to turn to Airbnb for space-sharing, or DriveNow for vehicle shares, than to other private individuals.
This is where trust really comes into the equation, as the more valuable the asset, the more likely we are to want to derive some additional income from it, while the less likely we are to trust a stranger to take care of it and not to steal it.
Despite the DAO debacle, slock.it have done some great work in this area, and their idea of smart locks, backed by smart contracts, are an easily understood way to communicate the benefits of blockchains such as Ethereum to the general public. Their Share&Charge initiative, where individuals allow their car charging stations, backed by smart contracts, to be used by other individuals, takes the idea a step further.
If a bicycle, a charging station, a garden shed or a lawnmower is able to take the initiative and hire itself out, maybe even paying for its own service and maintenance, we begin to view the idea of ownership in a different way. Possibly the company which makes the asset retains ownership, but humans are able to undertake some caretaker duty over the asset, earning themselves some percentage of the machine’s income.
It raises the tantalising question of how the governments of the future will legislate and even tax these transactions. Will physical assets have their own tax accounts, separate and distinct from the humans who derive income from them? In other words, will they become DAOs in their own right? Or organise themselves into DAO collectives?
As machine learning enables industrial robots, or autonomous cars, to adapt to their own surrounding and to transact with one another, will it free up humans to engage in more meaningful pursuits? Or make us collectively redundant?.
Devices that sell or share data
If the intersecting arcs of machine learning, IoT and blockchain can produce futuristic scenarios like the ones I’ve described, then the intersection of IoT, blockchain and machine learning has equivalent potential. Prediction markets like Augur or Gnosis become more accurate as they consume more data.
If data is the oil of the future, then it should be able to be traded in a perfect marketplace, to the entities which bid highest for it. Expect the cities of the future to be criss-crossed with networks of sensors, collectively producing petabytes of data, which some other device, wherever it is in the world, will discover and consume.
Want to equip your house or your car with sensors which monitor the weather, noise levels or pollution? Someone somewhere will want to buy that information. The use of smart contracts to enable markets — whether spread-betting companies or insurers — to purchase that information will be a growth area, as will decentralised oracle services.
But what about data that could be sensitive — how fast you drive your car, or the results of your recent medical tests or health data? This is where the need for trust comes in, and where there may be a case for blockchain in the strictest sense of the definition.
For individuals who wish to monetise their own data, or that of their connected devices, smart contracts are one way to define and limit the type of data that is disclosed to particular sources.
Similarly, devices which consume this data need to have some assurance of its provenance.
Take the example of the oil pipeline. Data is only as good as its source, and without some kind of third-party, trustless mechanism, it is worthless.
So, who are the companies and what are the technologies that will power this brave new future?
Whether the ultimate solution is something called a blockchain, something more like a blockless distributed ledger such as IOTA’s Tangle, or some other kind of peer-to-peer solution, the big question is whether these networks will be open and permissionless, or locked down by the existing corporate behemoths.
Bosch, BNY Mellon, Cisco and various other household names have teamed up with both Chronicled and Filament to participate in what looks increasingly like an R3 for IoT.
IBM has invested millions in blockchain tech (particularly in integrations with its Watson IoT platform), and is battling it out with the likes of Microsoft to offer blockchain as a service (BaaS) as the default for powering the industrial networks of the future. Microsoft offer Waves, Parity, Syscoin and Blockstack among the various blockchain integrations on the Azure Marketplace. IBM in particular seem to be aiming a charm offensive at would-be blockchain developers, making it easy to deploy projects using Hyperledger Fabric on BlueMix, with a free 30-day trial.
So, what exactly will the networked, blockchain-connected world of the future look like?
A future where our inanimate devices assume a digital identity and earn money for us, allowing us to participate in in an individualised, sharing society? Or a dystopian vision of networks controlled and connected by faceless corporations, sucking our data from us, sensing our presence, using our identity against us to deny us entry to particular areas, or to sell us things we don’t need?
Perhaps, according to William Gibson’s theory of unevenly distributed futures, it is already here.