 Good morning, everyone. I'm Bill Ren from Huawei. I'm glad to be here to share with you my thought about future network evolution. So the short video just now gave the title of my speech, accelerating the autonomous driving network, while industry collaboration. So talk about autonomous driving. So first, let's have a look at the history of automotive industry. So in 1885, the Benz Motor Wagon, the first car with a framework. In 1958, the Saab J750, the first car with a seat belt. And in 1978, the Cadillac Civil, the first car with an electronic control unit. And in 1999, Mercedes-Benz, the S-Class W220, the first car used adaptive cruise. So the first 100 years, the car industry improved in a continuous curve and focused on the traditional challenges, like the performance of the engine, the driver safety, and the fuel efficiency, and et cetera. But for the past 20 years, some new players started to pay attention from some new challenges, like global navigation, high definition maps, and identify the environment dynamically. So by last year, December, the WIMO-1 commercial usage stands for the autonomous driving has arrived. So by dealing with these new challenges, we clearly see a paradigm shift in architecture and ecosystem. So look at the history of Tahaka with an even longer history. So we also have a traditional challenge to deal with, like capacity, power consumption. But for the next 10 years, we have reason to ask ourselves, why not autonomous driving networks? Do we have a new challenge to deal with? Do we still do continuous curve improvement or paradigm shift? So maybe we should ask first, why autonomous driving network? So on one hand, the network complexity will continue to increase by the network scale will reach to 75 billion devices by 2025. And the SDNF technology will be deployed in a large scale. And the marketing space will be expected 100 billion. And on the other hand, the cost of the network operating and maintenance is huge. The efficiency of the carrier is 100 times lower than OTT players in some extent. And the growth of the all packs is much faster than the revenue in the past 10 years. And at the same time, the global digitalization or the transformation has all for huge opportunities to the network. The marketing space of SAT digitalization is expected 5 trillion. And the marketing space for manufacturer digitalization is 6.4 trillion. By 2029, the ratio of IT and the application move to cloud will be increased from 25% to 95%. All this means we have a new challenge to take care of. So like opening this internet driven and self healing, big data automation, all those requires paradigm shift in architecture and ecosystem. So first, let's look at the paradigm shift of architecture. A typical autonomous driving schema will be abstracted as a three-layered closed loop architecture. So at the bottom is a car, which needs to add a large number of sensors. In the middle is automation layer, which contains automation control plate form and big data analysis plate form. The result of the analysis will directly drive the control plate form to achieve this automation. And at the top is a cloud AI plate form, which contains the global data and have the AI model training and inference capabilities. We can see the architecture shift from product centric to platform centric. So similarly, for the autonomous driving network architecture, referring to this three-layered closed loop architecture, we should also add automation plate form and AI plate form above these network elements. So all those plate forms supposed to be public and open. So how to build this? Obviously, open source is the best way or effective way to build it. So referring to this architecture for autonomous driving network, the open source community should build a common vision and framework to drive the industry to work together. So like this acrimus for this cloud AI plate form, on-app for the automation plate form, on-app and Panda for this analysis plate form, and Acrena, Fido, OpenMFee to build and operating this optimized infrastructure. So we can see we have a full stack community to be leveraged. So refer to the maturity level of autonomous driving. Here, we'd like to give some definition and the details of autonomous driving network. So from bottom up, we can see the scenarios from deployment and service provisioning to network optimization, planning, and design. From left to right, step by step, all the manual work will be replaced by the AI robots to achieve from level two partially autonomous driving all the way to the level four, level five highly and fully autonomous driving experience. So compared to the car, most above level four already. So the network is still in level two today. So why is the network is lagging behind? So we have to talk about another paradigm shift. So this is a paradigm shift of ecosystem. While autonomous driving developed so fast, the rich ecosystem on the left gives a reason. So there are not only traditional car manufacturers transport operators, but also technology providers to the autonomous driving platform in the middle, in the middle. So many internet companies has provided those technologies needed for those automation platforms. So therefore, for the car industry, has entered a cross industry, open innovation, and collaboration mode from a single industry. So look at the ecosystem for the telco industry. Apart from operator and the windows, it's lack of technology providers for the automation platform in the middle. So there is no single company could offer all those technologies of the automation. So how to embrace the new technology and how to build this rich ecosystem? Again, the community of the open source maybe is an effective way. But the community and its pattern also need a paradigm shift. So firstly, the open source community need to do more than share the code and build a common WinF marketplace and also do the certification program. Secondly, in addition to the traditional operator and the windows, it's necessary to introduce some OTT players and the vertical industries to share the AI technology, software capabilities, and industry knowledges to achieve a better cloud and the network experience. So to accelerate the autonomous driving network, I call for Windows operators and the community to act. So for Windows, of course, it's better to add more scenarios and use keys to the community to define the new challenges and the architecture together. So Swisscom, Huawei, and Nokia build this broadband service orchestration with ONAP. So it's a good example. I encourage you to visit our booth for more detail for operators to speed up ONAP deployment to issue ONAP-related AFP. So Orange and Huawei has jointly verified this ONAP-based orchestration service with SD1 in your commercial environment. So welcome to you attend this track session tomorrow. And for communities, we should jointly build this ONAP-based market place. So Huawei has invested in a testing framework in OVP program. So I encourage you to visit our booth for more detail. So finally, as a summary, we have new challenges to deal with. We should deal with openness and intelligence beyond efficiency. So new paradigm shift to build this platform and ecosystem beyond the products. So new collaboration is needed, a joint vision and cross industry collaboration beyond a single community. So we'll find a thought about what matters in the journey of open source and in the past years, what we learned. So one, first, the open or closed is just a means. It doesn't matter at the end. It matters is an extremely experience you could bring to your end user. Second, the open is not an attitude. It's a capability. So it doesn't matter you like it or not. It doesn't matter you claim the open or not. It matters you could leverage it or not. So the third, I strongly believe we as an industry should work together and could have the capability to leverage the full stack open source community to achieve, to build, and to accelerate in this autonomous driving network. Thank you.