It's not surprising that Internet of Things (IoT) transactional settlement and data transfer layer IOTA is a co-founding partner of IoT consortium the Trusted Internet of Things Alliance.

As well as other blockchain IoT specialists like Filament and Chronicled, the new consortium features heavyweight industry players such as Bosch and Cisco.

IOTA's Tangle ledger is a Directed Acyclic Graph(DAG), as opposed to a linear blockchain design, allowing the system to settle transactions with zero fees enabling devices to trade exact amounts of resources on-demand, as well as store data from sensors and dataloggers securely and verified on the ledger.

IOTA founder David Sønstebø explained that big corporate entities such as Bosch and Cisco have tested the likes of Ethereum and are increasingly coming to the conclusion that it alone will not work for IoT; the next go-to ledger tends to be Hyperledger.

"But Hyperledger has its own limitations and its own problems," said Sønstebø. "While it might be good for enterprise solutions in the future, it's ultimately a permissioned ledger, and that's the antithesis of IoT, which has to be open in order to realise its true potential."

"It can't be a closed ecosystem because that's literally the opposite of interoperability. The problem we are seeing with Hyperledger is that it's very corporate driven; it's the typical old school way of doing things, which is fine for certain use cases, but not open ecosystems. The only IoT use case that they have experimented with, as far as I know, is the shipping one, where you track the supply chain for provenance etc - even those pilot projects have not really amounted to anything other than headlines. That being said there could of course be overlap between permissioned blockchains like Hyperledger and permissionless IOTA, as long as they are used in their correct domains."

Sønstebø pointed to three things that prevent traditional blockchains from being useful in the IoT world: fees, scalability and throughput, and the rigidity of infrastructure.

Because public blockchains like Bitcoin are essentially used by two types of participants – users and validators – there will always be an issue of transaction fees. Validators are only running their systems to earn fees, and as blockchains become popular this leads to bottlenecks and higher fees.

This is fatal for a situation that involves processing micro-transactions between machines. "If I'm a sensor and I want to purchase storage, analytics or bandwidth, those transactions will most likely be something like one cent or two cents. Imagine paying dollar-plus fees to send one cent – that's not a good business model," said Sønstebø.

"This also prohibits using the ledger for data integrity because the beauty of a distributed ledger is of course that you get this tamper-proofing – but if I have to pay dollars for each data packet and I have billions of data packets that I want to send, there is no way to put it on the blockchain in real time."

The second issue is scaling or throughput: 10 or 20 transactions per second wouldn't even really cover one connected vehicle, said Sønstebø. "When you have millions or billions of devices that are sending insane quantities of data and transactions, 20 transactions per second [tx/s] globally is not feasible for any kind of real world deployment."

The third problem is the very rigid infrastructure-intensive way that blockchain applications function, such as being continually connected to the internet at all times, in order to be synced to the chain.

Sønstebø said: "Even though the Internet of Things has the word 'internet' in it, in reality most of the connected devices will primarily use meshnet architectures locally and only relay to the main network (internet) perhaps once a day or twice a day in order to save battery. Meaning that if you use the blockchain you would end up on a fork and subsequently everything that happened in that partition would be invalid."

IOTA removed the decoupling associated with users and validators on a blockchain. Its Tangle architecture means there is only one type of party; in other words, the user is also the validator. And when a transaction is issued on the IOTA network, that user also validates two previous transactions. This allows the network to scale the more users are on it, while doing away with problem of transaction fees.

"So you have this emergent self-regulating, self-sustainable network. It's entirely participatory; you no longer outsource the validation of the network to some mining farm in China. And because there's no fees you can send 0.01 cent if you need that.

"We don't have blocks with fixed sizes either, instead it's a Directed Acyclic Graph, meaning that you can attach transactions where ever you want on this graph essentially, kind of in an asynchronous fashion.

"The more transactions the faster the network processes them. So it's the exact opposite of Blockchain, which becomes congested when there's a lot of users."

According to Sønstebø the network has already managed throughput of 400 tsx/s – but that's just the beginning. "There is no hard limit. We can do over 1000 tsx/s. That is not an issue as long as we have enough activity on the network. The real eventual limit in Tangle will be bandwidth itself, i.e. how fast radio waves or photons travel"

Regarding the rigidness of blockchain infrastructure requirements, IOTA adopts partition tolerance and eventual consistency in the CAP theorem sense, which means the Tangle is a lot more malleable. "If you issue a transaction in a cluster that's not connected to the internet but instead just communicates with other devices via some local protocol, then you can still retain the integrity of that cluster and then when you connect back into the main internet," said Sønstebø. "You seamlessly get this entangled kind of web, and that makes it a lot more versatile and agile in IoT ecosystem."