 Hello, good afternoon everyone. My name is Joran, the creator of FlowChain. Today my presentation will be about a whole new idea about the IoT blockchain. So I wish to speak slower and in a simple sentence. I try to make it easy to understand and also hope to gather feedback from all of you. So the FlowChain project is an open source hybrid blockchain for the IoT. So let me give you about five minutes instruction about FlowChain. So the project is simple. It's an IoT blockchain for the hardware and the first application of FlowChain is AI. Maybe I can surely demo in a bit later. So it's a whole new project. I mean a project was denied from the ground up. So basically we should have reviewed papers and open source projects. Especially the project is denied for the hardware, meaning that the FlowChain is a type of embedded system or operating system for IoT devices. And the project is in open technologies, means that we use open source licenses such as MIT, EBSD. We don't have any GPL in the source tree. So it's better for startups and companies to use FlowChain in their products. And especially the whole architecture is denied with old open standards. And we use the web technologies such as REST API, JSON objects, and JSON databases such as LabelUp, LabelDB, and MongoDB. Instantly the source tree is all in JavaScript. I want to ask you a question about this issue. Did you know why FlowChain was implement in JavaScript? The idea is to rate a FlowChain application a server runs on a variety range of the hardware devices from a high performance device such as the cloud servers and your attacks to a resource constrained devices. So you can see the finger. The finger shows that the FlowChain framework uses a Node.js Java to run time for a high performance device. While you use the JavaScript for resource constrained devices, it means that you don't have to modify the code to run FlowChain on different hardware devices. So the FlowChain system has its own architecture and consensus algorithms. Some details can be found in these reviewed papers. So the source code has already hooked on GitHub. There are so many repositories you can check it out there. So let me put a summary. So although there are already blockchain builds for fintech insurance appear to be payments and credit occurrences, there are still no blockchain denied specific for hardware. So FlowChain is blockchain technology denial level for IoT hardware with several features. But the overall idea is the then the distributed autonomous machines. So I didn't use the word organization. It's because for the FlowChain, I think all things is about data. How to generate trusted data with FETs for AI systems, such as the AI systems. It's a significant future from the beginning of FlowChain. And technically FlowChain is a private blockchain system for your needs. I will explain what's a blockchain private blockchain and public blockchain in a bit later. So open source is nature. If we want to build decentralized systems, it's because decentralized is impossible if we use the proprietary software. And if we have to use trusted startups to exchange private data, the decentralized model is also impossible. So as we know, this scenario is called trust is computing. And the blockchain is a fantastic idea to build such decentralized systems. So the FlowChain project aims to address such an issue by building blockchain for hardware enough for credit occurrences that they can trust each other without any trusted startups. This can increase the transaction rate among many current blockchain for Bitcoin. The transaction rate is about 50 minutes or even longer. And for Ethereum, although the time is shorter, but it's still about 10 seconds. But for IoT, the transaction rate is about many seconds. So FlowChain aims to address such issues as well. So how to build a dem, distributed tournaments machine? The idea is simple. We remove the trustee third parties. So the data flow can safely send through an untrusted channel. So which is called trust is communications. So that's the model. The FlowChain model. The under layer is the trustee hardware. We can use trust room technologies or Intel graphic processors. The Intel graphic processors is also used in hyperledger sawtooth projects for secure consensus systems. And the middleware is the trustee communications. The trustee means the users or code the IoT nodes don't have to get PKI, the public key infrastructure. They don't have to use public key and private key to secure their communications. So I will show you the hybrid FlowChain. It's a solution for IoT to trust this communication without PKI. FlowChain adopts the public blockchain. So for private blockchain, only 30 tech users can join the private blockchain network. But it's a model. The IoT nodes and IoT devices that execute the FlowChain application can be self-organized as a peer-to-peer network in a ring-top logic. By the famous P2P protocol code from MIT in 2021, I think. And you can see the architecture in the middle. The gray circle called hybrid node. The hybrid node is a device that receives the parcels from the public blockchain and delivers the parcels over the peer-to-peer network to the private blockchain. So the parcel means the miner. So in the public blockchain, you can see parcel miners layer. The parcel miner is a computer that aims to generate such powers and broadcast powers to the private blockchain. So I will explain the algorithm in a bit later. So in the private blockchain, all FlowChain nodes are running the FlowChain framework. The architecture comprised of three layers. The bottom layer is the WOT. And the middle is the broker server layer. Sorry. I don't want to explain the underlying layer. So I just put only one finger here for your interest. So the FlowChain adopts the WOT way of things on target, which provides a standardized model to represent physical devices as an IoT application server, which runs on the devices. Also, the WOT survey is a combination of connect and servers. So in short, this layer is supporting the peer-to-peer communications. FlowChain adopts the hybrid blockchain model. The hybrid blockchain system comprises of a public and private blockchain. One public blockchain with multiple private blockchain systems. It's because all mainstreams can safely send through an untrusted channels. Making leaks leads to trustless systems. So for a trustless system, it means that our ability to trust it does not depend on the intentions of any particular parties. So here is the algorithm. I understand that it's not very easy to understand the algorithms, but I put a slice here for your interest and wish to get feedback from you. Maybe not today at this moment, maybe by emails. So the underlying layer of FlowChain uses the code protocol to organize the devices in a ring topology. This topology is to reduce the connections and reduce the cost of the bandwidth. So how transactions go to FlowChain network here is a very simple demo. So you can see there are four chunk data. Chunk data means slice data. So like a live cam, a video camera, there are around 30 frames in a second. So the D1 means the first second and D2 means the second frame. So the first second is computed as a data key. And here the P2P network looks up the data key in a DHT, the distributed hash table. So for example, the code algorithm can look up the succession node to process the chunk data and then generate the transition IDs. I don't want to give more explanation about the algorithm, but there is a very important design here called the Kanomi code. Does anybody know about the Kanomi code? The Kanomi code is a magic string sent from the public blockchain to enable a fast authentication without PKI. So the system is called pseudo-assertication in a show. The distributed computing uses other four authentication techniques such as PKI to control access to the network. And most existing blockchain uses such PKI to authenticate users' secure communications, verify transactions, however, as time consumes. So some statics has finger lugs. There were several technologies that can repress the PKI such as pseudo-assertication. So frauds can use such authentication techniques to authenticate IoT nodes. So especially the IoT blockchain needs to authenticate nodes with a fast. And for Kanomi code labeled as a lambda here, it's a truly random string and can only use once. And it expires various strings like the app is Google Authenticator. So you may have used these apps before. So the puzzle in the public blockchain basically is a scheduler that provides a time-difficult Kanomi code. So the IoT node was authenticated to submit transactions by receiving the Kanomi code and solve the puzzles. The puzzle is a string search problem, not a crypto or hash problem. So I mean that the algorithm can be executed in IoT hardware devices, especially the resource constraint devices, which with limited memory and less computational power. The fully algorithm will appear in main. You can download the full white paper at fraudchains website. And also, it's an algorithm to submit a transaction to a public network. Also can be find in the white paper at the website. So my time is running also. I want to give a recap and summary in the shows. You can see in the table on the left side is the current trusty computing model. And at the right is the fraudchains idea. By using the trusty computing model, we can build an ecosystem of the dense distribution of autonomous machines. But the only problem there is that we have to use a trusty hardware. And currently there are only two technologies, such as Intel processors and trusty IP. So the future project is that things that come from Taiwan and the project was starting Taiwan. We want to work with the hardware industries in Taiwan or maybe in China. We are setting our office in China now to build a truly blockchain hardware chipset. So that's all. Here is my contest. You can check all the fraudchains website. The website is named fraudchains.co. It's very easy to catch up. And if you have any questions, please send me to your contact by search.