In its landmark Device Democracy report a couple of years ago, IBM highlighted a series of challenges that connected things would face when scaling from the tens of billions to the hundreds of billions in the coming years. Among these challenges, cost of connectivity, the erosion of trust in the internet, and broken business models for connected devices stood out as hurdles already affecting IoT innovation.
The report pointed out that using blockchain technology as the framework to facilitate transactions and coordination between devices would lead to a democratized “Internet of Decentralized, Autonomous Things.”
Since this report was published, innovation at the intersection of blockchain and connected devices has kept growing. While IBM developed its ADEPT concept, start-ups like Filament, Skuchain and Ledger use decentralized ledgers to connect industrial equipment or digitize supply chains. Last June, IOTA – developing a new framework for machine-to-machine transactions – launched its token, quickly reaching the top 10 of cryptocurrencies, and a market cap above $2 billion. IOTA’s innovation, the Tangle, seems to reflect IBM’s vision of a universal decentralized ledger. I spoke with Dominik Schiener, IOTA co-founder, to hear more.
Jules Besnainou: Can you tell us how you got to this intersection of IoT and Blockchain?
Dominik Schiener: The founding team at IOTA has been fascinated by IoT for a long time now, and in 2015 we started a stealth hardware start-up called Jinn. Jinn is developing a microprocessor dedicated to edge and fog computing environments, with the ability to start processing data locally. This microprocessor could enable a machine economy where devices start trading with each other. When we noticed the lack of existing decentralized solutions to settle transactions between such devices, we started working on one. That’s how the Tangle was born.
JB: Can you tell us what the Tangle is, and why current blockchain solutions could not fit your needs?
DS: The idea starts from the current technological limitations of blockchain. The limited number of transactions on public blockchains poses a scalability issue, and transaction fees are increasing, making them too costly for small transactions. The conclusion is that blockchain is not ready for production in enterprise use cases.
To tackle scalability and transaction fees, we came up with a completely new architecture specifically tailored for this IoT environment: the IOTA Tangle. It’s no longer blockchain (it has neither blocks nor a chain), but it’s still a distributed database and a P2P Network, and it relies on a consensus mechanism. Instead of a subset of the network being responsible for the consensus (miners), the entire network of active participants (i.e. devices making transactions) are involved in the approval of transactions. The consensus in IOTA is no longer decoupled from the transaction-making process: it’s an intrinsic part of it, and it’s what enables IOTA to scale without any transaction fees.
It is completely open source, its free to use, and we’re setting up a nonprofit foundation here in Germany to further develop the protocol and bring the stakeholders and large corporates together to further develop proof of concepts and case studies and to standardize this protocol.
JB: And what is your role in developing IOTA?
DS: A bit of everything. I do some technical stuff to help the the team, but mostly I’m helping in building the Foundation, this alliance that we’re shaping with corporates. I’m focused on developing partnerships, finding use cases, etc.
JB: Can you give us an example of such use cases you are developing?
DS: One of our first case studies is focused on a data marketplace for sensors, and we have a couple of pretty large corporates involved in that. The general idea is that any sensor is able to sell its data on the marketplace. As IoT is all about micropayments, we want to showcase what business models in this micropayment and machine economy context look like. It will function as a public demonstration, so anyone will be able to go online and purchase sensor data by paying a few cents.
JB: We look forward to trying that out! But how does IOTA make money without transaction fees? How do you compensate developers, for instance?
DS: The beauty of IOTA is that it has a different way of approaching consensus. Usually, you have miners and they have to be compensated, which results in a transaction fee. In our system, every active node in the network also actively validates transactions, and so none pay transaction fees. There is no reason to compensate participants. Obviously, because IOTA has very different incentive structures, your incentive to participate is to send money without any fees.
The way our developers are compensated is through the Foundation, which we’re setting up as an alliance where our corporate members pay yearly membership fees.
JB: So in a way, compensation is not based on how the system operates, but on how the Foundation caters to the general network?
DS: Yes, exactly.
JB: Could you explain how you guarantee security in a world without miners?
DS: Basically, the concept of miners is that a minor subset of the network is there to maintain the blockchain, the consensus over it. And we say that you don’t need a smaller subset, but you can make it possible for the entire network to participate and do transactions.
JB: So, does that mean that at a certain time, only the nodes that are transacting will be verifying transactions?
DS: Yes, so if you emit a transaction you have to validate two past transactions, and that is how the Tangle starts to grow in this graph instead of a sequential chain.
JB: So, the nodes that are not transacting at the moment will not have a role in validating past transactions?
DS: They still validate transactions, they’re just not attesting to that. Each network participant still maintains the ledger, the database at its core. And the database is validated by each node.
JB: What kind of role do corporate partners have in this Foundation?
DS: What’s really important for us, the hardcore technologists, is to get this feedback loop, to know how the protocol is being used, what kind of use cases are being explored, what kind of proof of concepts and what kind of limitations are being seen by these corporates. Then, we can start improving the protocol and the software itself, and fulfill our main goal to bring the blockchain to this production-ready phase.
JB: What about connectivity costs? You still need to get plugged to a network to get to transact with other devices, right?
DS: Yes, and networking is the biggest bottleneck. It’s the most difficult component.
JB: What kind of networking are you using at the moment?
DS: IOTA is partition tolerant, which means that we don’t need one ubiquitous networking protocol, we can have clusters, mesh networks. And as I’ve said, IOTA was created specifically for IoT, and so now we’ll aim to be network-agnostic.
JB: Does that mean that in the future you will be able to have devices that use Sigfox and devices on 4G, for instance?
DS: Exactly, or Bluetooth, or many others.
JB: Okay. And when do you think that will be possible?
DS: This summer. We’re currently re-architecting the implementation to make it very modular.
JB: And if I’m looking at this from the device standpoint, the device will have a processor maybe from your hardware start-up, maybe another one, it will be linked to a network and perform its usual functions. So if it’s a microwave, it will transmit data based on it, then the IOTA protocol is there to communicate with other machines, right?
DS: Yes. IOTA is two things, really. It’s there to do this transaction settlement layer, and it’s there for the security of things, in terms of data integrity.
JB: Okay. And how does this second part work?
DS: You can utilize IOTA without having tokens. With every other ledger you need tokens to be able to participate. In IOTA, you can make a transaction that either contains money or data, and if you do it with data you can obviously secure the data, so you have this data integrity component.
JB: And as you don’t have transaction fees, it’s free to send data around, right?
JB: Is it a good guess that data is going to represent most of your transactions?
DS: I believe that initially data is going to be the main use case, simply because other uses cases like machine-to-machine payments are a bit further away, but we’re already working on those as well.
JB: Right. But if we take the example of a self-driving car transacting with an EV charging station, the idea is already there and it’s great to have the architecture for it, but the actual market is quite far away right?
DS: That’s true. It will take time.
JB: Do you see use cases for machine-to-machine value transactions today?
DS: It’s a good question. Something like parking or tp;;s, or those use cases with direct value-added services for humans where it could make our life a lot easier would make sense in the near-term.
JB: So, you mentioned this use case about the data marketplace, do you have any other pilots that you can tell us more about?
DS: Yes. There is one in the healthcare sector, to secure data from IoT health devices. We recently set up a workshop in Norway with Oslo Cancer Cluster. And another one we’re working on is the ‘Carpass Project’ together with Innogy and Volkswagen, and this project is focused on securing the data inside vehicles and giving the vehicle an identity in the future.
JB: Can you expand a little on the idea of scalability?
DS: The beauty of IOTA is that its horizontally scalable. The more nodes you add to the network, the more scalable it gets simply because you can get more transactions through, and they get confirmed faster. So, IOTA when it comes to confirmation is not done sequentially as with a blockchain, but asynchronously so we can parallelize the confirmations.
JB: So, what you’re saying is that actually the more nodes you have, the faster you can confirm transactions, is that correct?
JB: And how do you deal with situations where there aren’t that many nodes?
DS: That is actually the current environment. We have a controller, a simple coordinator, that helps confirm the transactions. The main assumption behind the Tangle is that it is a large enough network to be secure and we have to get there first, and right now we’re in the bootstrapping phase.
JB: Can you give me an idea of numbers – at what number of devices it’s starting to be really robust?
DS: That is something we are currently exploring with simulations. We have access to two supercomputers with which we are running simulations. And we have three mathematicians on board who help us with this question.
We currently have several hundred participating nodes, and it is strong but not yet secure. It cannot run on its own yet. For that we need higher transaction throughput, which we’ll know more about with the supercomputers.
JB: In the current IoT world, how would you define the incumbent solutions? What are you replacing today?
DS: I don’t think we’ll be replacing anyone in IoT. It’s a completely new field for machines to transact with each other. I think Visa and MasterCard are the biggest competitors because they’re also trying to get into that IoT space but they’re completely centralized, and they have transaction fees, ongoing costs, etc.
There is no real M2M transaction today that can be taken seriously. Most of the time it’s still done centrally, over the cloud, where the machine simply triggers a request to a server and the server then does the payment, not the machine directly.
JB: Can you tell me a little bit about the team’s size today and maybe your goals in the next year or two?
DS: In total, it’s 22 people of which 13 are full time and the rest are focused on BD, case studies and ecosystem development, so not just hardcore techies. Over the next few years we want to introduce other protocols as well. Right now, we’re developing one for smart contracts, specifically for IoT. And over the next years, we also want to build up a Foundation with a governance structure that corporates feel comfortable in joining. Through that, we want to establish a European Alliance, a blockchain alliance or distributed alliance that is run in Europe that many of the large corporates are involved in.
JB: Thanks, Dominik. We look forward to hearing more news from IOTA!
DS: Thank you!
Dominik Schiener will be speaking at our next executive summit on Blockchain in Energy and Industry, taking place on November 9th in Boston. Read more and request an invitation here.