Artificial intelligence and robots have already talked about how they can no longer be buzzing, allowing outsiders to look like flowers in the fog. So, what is the opinion of the academic community and the big brothers in the industry regarding the more practical problems of robot application and development? The CCF-GAIR Global Artificial Intelligence and Robotics Summit held the first round table forum for robots and drones. Host "computer" Creation and former president, a former chairman of Mr. Chen Zongzhou SAN FRANCISCO, Professor of Computer Science and Engineering, University of Southern Florida, a visiting professor at Stanford University, Sun Yu, Chen Shikai Si Lan Technology, MIT robot experiment The chairperson, IEEE, AAAI Fellow, and Daniela Rus, academician of the National Academy of Engineering, participated in the discussion.

The key to the robot's future life: low-cost and cerebellar research Moderator: First, Mr. Chen Shikai asked us to give us three key words. One is the future, the other is life, and you talk about what you know.

Chen Shikai : I know that the theme of today's roundtable is the robot's future life. Then these three keywords are very appropriate in the theme. Because everyone is thinking about the future of human life, there will be no shortage of robots.

In fact, why did you set up such a company? When we began to imagine the future of human life, we envisioned: How do we let robots in human life, on the one hand undertake their own work, and on the other hand can live in harmony with humans?

Regarding the future and life, we expect that things that are tedious in life, such as household chores, caregivers for the elderly and children, and sending and receiving courier for online shopping, can be solved through robots, allowing people more time to enjoy life. Life becomes easier and more convenient. In the process, in fact, the future of our imagination may be relatively distant, and it may even be out of reach. However, we hope that what will happen from many years after people think about it will eventually become tomorrow, so the future The life needs to be solved through the innovations of existing technologies in the academic and industrial worlds to solve the problem that we still have at this moment in time for future robots to play more roles in life.

The professor also introduced a lot of field exhibitions on the front of robots, such as flexible robots and collaboration. Of course, there are still a lot of problems that need to be solved in the industry, such as self-help actions or autonomous positioning and navigation of robots solved by Sci-Tech. Because we know that the main difference between robots is the ability to control them autonomously. Unlike humans, robots do not have such capabilities. In the final analysis, it is necessary for academics and industry to work together to achieve this function.

At present, the service robot industry is still in a relatively early stage of growth and needs a corresponding growth period. Autonomous navigation is also very important for robots, but cost is the biggest obstacle at present, such as many current third-party laser radars. Such an environment is not conducive to putting advanced technologies in the industry and pushing them into the consumer market. This has affected the development process of domestic robots. Therefore, our goal is to reduce it from the original price range of RMB 10,000 yuan, through the independent research and development of new technologies to the promotion of tens of thousands of households and to build a standard of future life.

On the other hand, robots have corresponding eyes, brains, and very important parts of the cerebellum. Human-related movement regulation is carried out by the cerebellum. For robots, not only need a very smart brain to communicate with people, or interact with the environment, but also must support the cerebellum of autonomous behavior and self-help actions . Including Google driverlessness is the reason for this, and cerebellar R&D is also an area that has been continuously explored since the establishment of Sihan. Because after Lidar, we used it as a standard modularization and developed some interfaces that could be used by the outside world to expand the use of robots, allowing robots to have more eye movement after they had eyes. In this way, our fantastic future will no longer be so far away. This is the direction of our efforts.

How does the frontier artificial intelligence laboratory "engage in production"?

Moderator: Thank you, Mr. Chen Shikai. Next, I wanted to ask Mr. Sun Yu to introduce his work, but he will share with you later. Then he wanted to ask the two guests a question.

Sun Yu: Let me first pose a question to Daniela Rus. The MIT Lab has many talents and talents. You are the person in charge of the lab. Many of them are interdisciplinary engineers and researchers. How do you get them to collaborate on a project?

DanielaRus: Actually, we don't have any magic bullets. It's true that our lab has passionate and talented people. This is very important for organizations. Any successful organization must have talent. First of all, time is limited. How can we make time fully utilized? Our laboratory is about 1,000 people in size, including 150 colleges. The teachers hope to improve their computing power and improve their ability to handle and mine big data. At the same time, some people will love Internet games. Some people want to be robots. Everyone has their own ideals and wants to do things. But in our labs, interdisciplinary teams really need to let them cooperate.

How do we do it? We have adopted different projects and initiatives. These projects include the cooperation of academia and industry, such as big data, and cyber security. We hope that talents in the academic community will actively participate in the industry and encourage researchers to work hard to get their research results to the ground.

Which computer vision technology is reliable?

Moderator: Thank you, Daniela Rus. So Professor Sun Yu, your second question is to ask Mr. Chen?

Sun Yu: Lidar you do is of course very good, but now there are two (laser radar). Why did you choose this?

Chen Shikai: Actually, this issue is also an issue that has been discussed in the industry. That is why we mainly consider several aspects:

First, at this stage, especially in the past two years, the research on camera or vision in the academia is making rapid progress, but it has encountered many problems in the industry. First of all, this technology requires very high computational requirements.

The second aspect is the popularity of robots. There are still many obstacles in the more advanced functions of the academic community. In addition, changes in the scene are also very challenging. The human eye has a higher sense of sensitivity or sensitivity than ours. The camera may be a trend, but in the industry not only is it necessary to take advanced technology, but it is also more practical to observe the most mature technologies in these two years. The industry has always been practical, and the history of laser research in academia has been long. Currently, academic circles are already very mature technologies, but there are still many difficulties in the industry. In general, from the perspective of what we want to do, laser is a very reliable technology because the cost is solved.

In fact, we believe that convergence is the future direction, combining laser, vision and other technologies to generate converged solutions.

What is the progress of robot production costs? Moderator: Because robots have entered the issue of life, the insurmountable barrier is the cost issue. I would like to ask Prof. Sun Yu, now with the hands, eyes, and the whole part, what are their progress in terms of cost?

Sun Yu: This is a very good question. Robotics started in academia. In the academic world we can count on the cost. But in practice, even if you do not want to do housework, you are not willing to spend millions to buy a robot.

Therefore, in the last decade or so, in addition to industry and academia, the academic community has begun to think about the issue of cost. We hope to push the robot to millions of families. In this regard, we did a lot of things: First, simplify the structure of the robot, and another is to try to reduce the cost of the motor. Microsoft's cost is $100, so this problem is well solved. From all aspects of robot design and visual angle, we are working toward these aspects.

Auto Power Transformer

For the Oil Immersed Aouto Transformer, we can produce voltage of 220kV, 330kV and 500kV and capacity upto 460MVA. We use the best quality of raw material and advance design software to provide low noise, low losses, low partial discharge and high short-circuit impedance for Power Transformer.

Our power transformer are widely used in national grid, city grid, rural grid, power plant, industrial and mining enterprise, and petrochemical industry.

Power Plant Transformer,Auto Power Transformer,Auto Transformer,High Quality Auto Power Transformer

Hangzhou Qiantang River Electric Group Co., Ltd.(QRE) , https://www.qretransformer.com