Thunderbolt 3 Cables,Thunderbolt 3 Cable,Fast Charging Thunderbolt 3 Cables,Usb C To Type-C Data Cable Dongguan Pinji Electronic Technology Limited , https://www.iquaxusb4cable.com
NetEase Smart News reported on August 25th that the 2017 World Robot Conference was recently held in Beijing. During the conference, Yu Haibin, Director of the Shenyang Institute of Automation at the Chinese Academy of Sciences and Director of the National Key Laboratory of Robotics, delivered a speech titled "The Robot Internet." He discussed five critical issues concerning the development of the robotics industry, including where the future development of robots is headed. Against this backdrop, where should our focus lie regarding work priorities? What are the existing challenges? And where will future efforts be directed?
Below are excerpts from Director Yu Haibin's lecture:
Yu Haibin: I recall that last year’s presentation focused on underwater robots. Since everything involves the physical world, it’s better to discuss practical applications. Today, we’ll continue exploring cutting-edge topics. I’d like to share some thoughts on what I imagine and will pursue in the future. Hence, my report is titled “The Robot Internet.â€
My presentation will roughly cover five areas: First, where is the future development of robots heading? Second, under this context, where should our work focus be? Third, what are the current challenges? Fourth, where will future efforts be directed? Since discussing this broadly would be too extensive, we’ll delve into specific fields.
The development of robots has seen exponential growth. The core essence, as stated in McKinsey’s 2013 report, is more clearly defined, focusing on both economic and technological perspectives. At the time, the most striking takeaway was the mention of a "trillion-dollar market," which was unimaginable in the robotics field previously. Later, we thoroughly examined this report. Why did the robotics industry suddenly expand to such a vast market? We discovered that the key lies in the constant expansion of the field. Initially, robots were primarily associated with manufacturing, particularly industrial robots. However, this report also covered service robots in areas such as healthcare, wellness, and defense.
In this context, it becomes evident that robot technology requires new breakthroughs. Indeed, many new concepts have emerged, sounding similar to traditional robotics but incorporating elements like sensing, intelligent software, human-computer interaction, and networking. My impression is that this is the first time we’ve collectively emphasized the holistic requirements of network interaction software or sensing. Previously, programming for viewing and teaching was more common, whereas now there’s a greater emphasis on automatic programming within task sets. We’ve discussed integrating networks with robots, but actual interactions remain scarce. Many people merely copy concepts and technologies from the information sector. A robot also needs an operating system. What exactly is an operating system? Essentially, it’s an abstraction of physical ontology. Thus, the entire report reflects the entry of information technology. It’s also apparent that if service robots perform well, it signifies their potential to transition from structured to unstructured environments. This is good news, indicating that information technology should play a crucial role.
Currently, we face numerous challenges, but overall, considering robots as terminals, the mobile phone remains the most successful example, revolutionizing certain concepts. Initially, we had the distinction between IT and CT, but now we have ICT. It’s hard to categorize a mobile phone as solely an IT or CT product—it’s a comprehensive entity. Similarly, for robots to reach such a level or achieve a trillion-dollar market, they must permeate every aspect of our lives. Robots are more complex to use. Generally, professionals are still at the stage of programming. Integration also presents challenges. If robots move to a new environment, integrating well with the existing system remains a significant hurdle. They must seamlessly link with infrastructure wherever they go. Adaptability poses another challenge. Robots still have considerable room for improving their understanding of environmental and non-structured settings. Consequently, we face challenges in robot programming, environmental awareness, and human-computer interaction.
How should we proceed? We have multiple solutions, including some disruptive approaches. For instance, advancements in new materials, biotechnology, and brain-like intelligence might yield progress, but in the era of electromechanical systems and system integration, we believe the integration of network information technology and its integration is the most realistic path we can envision. For example, when discussing self-driving cars, some see them as information terminals and part of intelligence. Observing the periphery, the information society has entered the era of the Internet of Everything. What is the concept of the Internet of Everything? Globally, there are currently 7.6 billion people. By 2020, the massive use of IoT terminals will reach a scale of 50 billion units, representing a two-order-of-magnitude improvement. This progress brings us many challenges. Information professionals realize that infrastructure is a significant challenge.
Where do these 50 billion terminals come from? Who are the most advanced terminals? Who are the terminators of the missions? Nowadays, some typical terminals and successful enterprises, such as Huawei’s well-developed mobile phones, are essential. Besides these backend networks, the most successful information terminal remains the mobile phone. Why can’t robots achieve the same? They should, and I believe they must. Many of the information terminals we see today, including cameras, are relatively simple yet intelligent. The intelligence of robots themselves is sufficient, but how can they become popular Internet terminals? In fact, this is also the most likely direction to address the challenges I mentioned earlier.
This is a guide from the U.S. National Science Foundation last year. Most discussions revolve around collaboration. Here’s an example. What does this collaboration depend on now? Some people have proposed ideas. Let’s see if there’s more room for improvement in these areas. Of course, these things are just beginning to emerge, as we’re discussing cutting-edge technology to clarify these matters. Now that there’s more talk about cloud robots, I think the industry has differing opinions. Some robots think they are engaged in this, while others think it’s more about information. Either way, it’s a sign.
Looking back at the development of robots over the past 50 years, the first industrial robot was soldered to the industrial sector and has since moved into large-scale manufacturing and electronics manufacturing. Then it diversified into the manufacturing industry and re-entered the household. It then exploded and grew exponentially. Regardless, the examples given here are clichés of robots that are not connected to infrastructure. Let’s consider what can be done to transform human society in the information age? If we can make it known to every family but no one knows about it unless it connects with infrastructure? Cars are everywhere, supported by a robust infrastructure. Mobile phones are ubiquitous now, with strong network infrastructure support.
Cloud-based robots indeed have their advantages. The general concept of cloud-based robots is how to utilize the cloud’s computing power to enhance robots’ deficiencies in various aspects. For example, when we talked about computers, we mentioned client-server models, computing grids, and even thin clients and thick servers. The cloud provides immense computing power. We believe that many smart functions in the future will be implemented on the cloud. This is most likely. The cloud’s computing power is powerful enough. This raises a good question: how do we combine it with a robot? Of course, there are many challenges. Now that we’ve just begun, we can’t rely solely on information. We have a lot of room for describing and digitally integrating robots. So now we have many CPS concepts. It’s just a concept. There are still many problems that need to be solved.
The RoboUs project is a Dutch initiative. The fundamental goal of this project is how to fully utilize the capabilities of cloud computing to solve the "intelligence problem" of robots and shift heavy computation tasks here. It does help with some of the basic problems of robots, such as SLAM, providing some advantages. Regardless of size, the core is to use the cloud’s ability to store and share information. It’s the idea of offloading computation to the cloud and solving some of the collaboration problems. So there are many technologies to develop. The bottom corner is more about what technologies are being worked on. I think this is a good idea, but what does it solve in terms of interconnection? This is vertical interconnection. It’s a connection from the object side to the cloud side. It has its advantages, but it doesn’t solve all the problems.
We believe that the robot internet is a system. If there are people down there, we are reviewing the history of the Internet. This is probably the case. The core of the Internet is no longer just about technology. Behind it are decades of development and progress. For example, the core behind it is the development of integrated circuits. Without such an integrated circuit, today’s Internet wouldn’t exist. On the other hand, the Internet is an advancement in network technology. If there’s no IP technology, the Internet couldn’t be so easily extended, allowing us to access everywhere. However, we will see the Internet go further. We won’t talk about the recent middle process. Many content services have already mentioned this, and the core of the Internet of Things has erupted for the second time. One of the core advances of the interactive Internet is cloud computing because it finally gave us a platform that can be calculated. It turned out to be just a communication channel.
Let’s talk about some things of the Industrial Internet. We must not completely cover up the true meaning of the concept with this term. Strictly speaking, the industrial Internet is not just a network. We shouldn’t think that the Internet used in industry is called the industrial Internet. It’s an integrated integration of two aspects: On one hand, it promotes the use of network technology in industry, and on the other hand, it makes industrial applications based on these networks. Because this is first proposed by GE, it proposes the integration of the three elements of people, machines, and data in manufacturing. From this perspective, the Industrial Internet is not just a network but a system. Everyone says this is not a bit boring? Actually, this is not exactly the case. Under the advocacy of GE, IIC has now become a coalition, and many people in the industrial world are leading its needs. Information technology is integrated. The concept of integration is not simply used. It’s actually very much about people who have been engaged in automation for a long time using the results of information technology and achieving good fusion. Now that we have developed into the 5G era, we say that CPS is an end-to-end development of information. Everyone should be a common process, not a simple computer. To give a simple example, if cloud computing technology is only used for robots, we can solve many problems, but many things are not suitable for you to do, so the more we rely on the end of things, we should start working collaboratively at the time.
Just talked about Internet robots, this one step closer than the cloud robot. Because robots and robots have been interconnected for a long time in the 1990s, but at that time, they used existing network technologies instead of making their networks based on the needs of robot collaboration. If we look at this from this perspective, there is still a lot of room for network advancement, so we have discussions with the 5G Forum. They also need such needs. The Internet of robots is based on the Internet robots. This issue is raised from the perspective of robot application and promotion or technology requirements. So the core problem is that there are two interconnections: one is horizontal interconnection and the other is vertical interconnection. The horizontal interconnection is the cooperation between the robot and the robot. The difference between the machine and the robot is very big. The robot can move and autonomously recognize the environment. The coordination accuracy between several axes when the machine is connected to the network is very high. But after all, it is under a radius. If robots want to collaborate in an area or in an unknown environment that is mobile, they must be wireless. The real-time nature of the robots cannot meet the requirements at all, because the two collaborating guarantee the accuracy can make a definite trajectory.
We propose the concept of robot Internet. Its connotation should be the integration of machine, human, physical process, information and human beings. The other concept is to have a support platform. We are based on existing technologies, but we must renovate and upgrade existing technologies, especially since 5G has already arrived at the physical end. The biggest difference between 5G and 4G is that it has customized many of its own standards and requirements for each area. Therefore, a new breakthrough should be made in the field of manufacturing, especially the field of smart space that can be used and can be widely used in the future. If only from the point of view of communication, we cannot solve it with many problems. So the robot Internet is a new model and a new application.
Connotation is valuable to developers and extension is valuable to users. Therefore, from the perspective of the network, robots are used to promote the deep integration of information physics. From the robot’s point of view, it is the ability to use the resources of the Internet to expand itself. We focus on one vertical and one horizontal. Then it should have a framework. We can see that this architecture has orange, which is the edge computing resources. At present, the computing resources of the Internet are large enough, but operational robots require real-time performance. Real-time performance is impossible if they all come to the cloud. If we do it at the end of the world, because we have to move it must be an embedded system, if it is local it has limited computing power. The end of the calculation to achieve three seamless: The first is the real-time communication can not be guaranteed, the gap to be resolved. The second is that the ability to calculate resources is not enough, there must be a set of information scheduling, because it should be transparent to the user. The third is that these business relationships are not well described, and many still require human intervention. These three gaps need to be resolved.
From the solution point of view, the robot’s ontology must have a good information abstraction and description. It’s not a simple controller concept. Only in this way can we add a lot of good things, such as information security and public safety, to provide a good infrastructure for robots. In the aspect of the Internet, in particular, the real-time nature of the Internet of Things does have many challenges. For users, it must be a transparent programming environment, and there must be a common architecture for cloud storage and programming. For example, another example is Google’s Internet crawling training. The 14 robots are all in a coordinated manner. After a lot of training and deep learning, they have strong intelligent crawling capabilities in complex environments. But now I’ll give you a question. If the robot technology is mature, how can it be applied? The current development workload is very large and there is no good platform. Google’s search engine is very important, so we will continue to be exposed to such a concept as the engine platform and basic services.
The challenges are highlighted in several aspects: Because of the seamless integration, basic technologies are challenging, and IoT has serious challenges, especially discrete manufacturing is a high-speed process. Such an indicator is difficult to achieve now. Even if such an indicator is met, it cannot meet the requirements of group robots. There is still a lot of work to do. Now, discrete manufacturing is the target to get high-speed and high-density robots. The environment does not fully meet the requirements. The computational challenges Just as I said just now, it is difficult to meet the real-time requirements for the migration of computing tasks to the cloud. Local resource constrained processing cannot handle so many complicated calculations. Because of the integration of sensors and information, it is not easy to find a solution. Location can solve the problem of time. We introduce some related developments. There are also some scattered work. In particular, the end-to-end network aspect has basically reached the point where 100-node network nodes do not need to be interconnected through intermediate nodes in the case of connections. The delay is less than 10 milliseconds. These are problems that can be solved without extreme high-speed movement.
The AGV control solution in the workshop can solve the problem of autonomous scheduling and problem-solving of 100 AGVs through this network. Because the current methods are partitioned, and more is done by centralized control at the top. Some integrated development environments have also made some adjustments. The original is more to the physical system. Now it abstracts a good development environment, so that you don’t need too much understanding of users and cognition. Users are open. We also need to make a customized production system, relying on such a simple progress. Because the platform is not yet available, some interoperable tools have also been developed to enable rapid industrial development across multiple robots. This is a production line that we made a demonstration and achieved some good results. Both the exhibition in Hanover and the World Internet Conference in Wuzhen have caused some repercussions.
This is a very challenging job. Our roadshow has just begun. But since GE talked about industrial manufacturing, it raised the industrial Internet and there are so many problems to challenge. We can certainly learn from it. As a robot, we must be able to reach the future more widely. We still have a lot of challenges in infrastructure and basic platforms. Of course, the background of this platform also contains a lot of basic problems. In short, we need to integrate deeply with information technology. However, robots have their own characteristics. We must make some contribution to this matter of the robot and we also hope that these can have a little inspiration for everyone’s future study and work.