On March 27, 2018, the “Link Future – 2018 Spring Blockchain Technical Forum”, sponsored by the Blockchain Section of Peoplecapital of people.com.cn, was held in the New Media Building of People’s Daily. Roy Li, the founder of Ruff Chain, was invited to attend it and, as an honorable guest, share at the forum his opinions about the use of blockchain under IoT scenarios.

Even before the forum, Roy Li mentioned that, in the world of blockchain, cognition upgrading is the most essential. Although he is the on-duty group leader of the 3 o’clock group today, he is more willing to popularize the knowledge of blockchain to beginners through the activity of Peoplecapital.

Concerning the current application practice of blockchain in IoT, Roy Li expounded his views in the following three aspects:

 Not all data are suitable for going onto the chain

Unlike the cheerful view outside that “the blockchain can change the world”, Roy Li believes that, not all the data in IoT are suitable for going onto the chain. Despite the highly attractive ideal of data non-manipulation, the distributed databases of blockchain can not bear the costs of storage and flow of huge information data.

“What has somebody said, or messages somebody has sent – such data are not suitable for the blockchain”. Roy Li explained that, the data of traditional internet products can be further divided into two categories: the information network at the presentation layer, and the value network at the back end. Although the data at the presentation layer account for 99%, or even more, the data in the value network are more likely to generate value in IoT.

The popular Mobike, for instance, belongs to the wide area IoT; by contrast, the PV power station to which Ruff is now being connected belongs to a local area IoT, and their data information belongs to the value network. And it is of great value to guarantee the authenticity and non-manipulation of those data. Take articles weighing in logistics for example. It is highly likely that, such manually-measured and recorded information may deviate due to personal status of work and spirits.

Therefore, it is adopted in Ruff Chain that, unmanned operation modes are applied at the terminal, and edge nodes are utilized to acquire data; then, the other data that are with the highest value and can be used for value flow and value circulation and whose consistence shall be absolutely guaranteed are sent onto the chain.

“The essence of blockchain is not to improve the productivity, but to improve the relation of production, to achieve the non-manipulation of data”. Roy Li then went on to explain the solution to “double-spending” and data silos with the blockchain.

 The selection of blockchain consensus algorithm is up to the application scenario

At the enterprise-level application, we often face such problems as “which algorithm shall we use on earth?”. Currently, a great many blockchain technologies involve the selection of consensus algorithm.

“This also depends on the application scenario we face. If you face a scenario with very low centralization, you should ensure sufficiently strong security. At first, I suggest you not use PDBFT; it seems that the fault tolerance of Byzantine Fault Tolerance is strong, but it doesn't work well with too many nodes, because it is directly proportional to the square of node quantity.” Roy Li expressed his view on the selection of consensus algorithm.

Also, he further explained that, the use of DPOS as the consensus algorithm of Ruff Chain does not mean DPOS is the most secure; it is only because DPOS is the most appropriate for IoT scenarios.

Therefore, the selection of consensus algorithm is based on your application scenario. Sufficient security shall be guaranteed if there is high requirement of consistence. In the field of finance, for example, when problems are solved in a centralized mode, dedicated hardware is often adopted for strong isolation; one hardware is used for one application. Thus, so far, the most powerful point-to-point payment system is still bitcoin; I'm quite sure that Ethereum is not as secure as bitcoin. The ultimately critical point is that, which point is your application based on.

 The core of IoT is how to abstract objects

In real IoT scenarios, the equipment we face is over a thousand manufacturers; each manufacturer has data of a variety of model, and these data are all different. Therefore, how to abstract these objects and how to abstract the data you need are the issues to be addressed at the application layer of IoT.

Data at the IoT end tend to be at the back end, so it is vital to be compatible with different equipment and data structures. It takes Ruff 4 years to be capable of being compatible with a large number of equipment on the market. It’s really a huge project.

Thus, there is a hardware abstraction layer above the Ruff Chain, which may abstract all the hardware, so that it is unnecessary to store a semaphore of high potential or low potential when such data are sent onto the chain. Besides, one more thing is done above the Ruff Chain: we turn them into light nodes based on these nodes. Since the storage space of embedded equipment is not big, it may fail to recall even in case of a status; these equipment can be verified, and set up in the middle.

“The OS of Ruff has ensured the acquisition of edge data and ensured no human intervention, so that relatively clean data are sent on to the chain and corresponding interfaces are provided, capable of data interaction with other chains”, explained Roy Li.

When more and more industrial IoT equipment and devices are connected in Ruff, both factoring business and supply chain finance can realize asset-backed securities based on the on-chain data, to issue ABS.

“The concept of traditional production company is more classical than that of classical internet. Perhaps, it is time now to help traditional enterprises with IoT, through the blockchain technology, to realize corner overtaking”, ——Roy Li