There are generic fenders for bicycles that can be fitted to most bicycle frames. These catch and redirect road spray thrown up by the tires, allowing the rider to remain clean. They are increasingly common on bicycles in the more car dependent English-speaking countries, as bikes in these countries tended to be SPORTS-oriented, with minimal clearance between tires and frame tubes for fenders, or were only ridden in clement conditions, although there are a few fenders that are designed to be attached to the Seat Post. The trend towards urban biking which has doubled in the past decade, lead to production, import and use boom in utility bikes. European utility bicycles, traditional roadsters used in developing countries, serious touring bicycles, and beach cruisers are nearly always fitted with fenders, universally known in British English as mudguards. They are also fitted to motorcycles. Bike Fenders, Plastic Bicycle Fender, Mountain Bike Fenders, Sks Bike Fenders Hangzhou Datasports Equipment Co.,Ltd(Oembicycle) , https://www.oembicycle.com
"Moore's Law" in Robotics
Maybe everyone thinks this picture is futurism, but in fact we have achieved it to some extent. Robots can be used to send parcels, clean up the environment, organize goods, autopilot, and life assistance. In addition, we have also seen companies that have invented two types of single-armed robots and have applied them to production.
What these examples tell us is that robots have indeed changed from science fiction to current scientific reality. We can make robots more capable and more intelligent.
It is worth mentioning that there is also a subversive law similar to "Moore's Law" in the robotics field. Including manufacturing tools, design tools, and other areas, the subversion of Internet performance will change every six years. Similarly, the number of factory robots will double every five years. At present, we have already verified this fact for the time being, and I believe that the frequency of this subversion will be higher in the future.
In the future world, everyone may have a robot. The robot is just as common as a car running on the road. I call it the world of ubiquitous robots.
These robots will be able to work with humans to accomplish many tasks. Of course, we have not yet reached this stage because there are still many technical issues that need to be resolved. For example, how do robots interact with people, how do they reason and solve problems by themselves... And how can we create new robots quickly and inexpensively?
Next, I share with you some technical trends that will help us solve the above problems.
Twelve major robot technology trends
Software robot
The previous robots were all steel bodies, but such structures were not well adapted to various environments. Software refers to the robot's structure is made flexible and flexible, like the human body structure. In general, the body structure of a soft body robot is made of soft silicone to enhance its adaptability and adapt to different unknown environments.
Based on the principle of muscle operation, we have found that such a configuration makes the robot more agile and able to complete certain tasks more quickly. In addition to soft silicone, we can also use water or air to drive the structure of the software. For example, in this (enlarged version) robot, it looks like a snake, and the surface of these bubbles can drive the robot's activities by zooming in and out.
We can see that when the robot is placed in the pipeline, it can automatically detect the surrounding environment. The adaptability of the plastic type is unmatched by the steel robot.
Similarly, we can also create robotic fish. It can behave like an actual fish, can make a 90-degree turn, and can quickly escape the hunter. Thanks to its soft tail, robotic fish can move up and down in the water.
We have already seen the importance of a soft body, and a new journal "Software Robotics" has been published for two years. Through this journal, we know that robots are ranked highest in importance in robotics, which means that everyone pays the highest attention to robots.
Manipopulatetion: Flexible Operation
In addition to software robots, another technology to improve and improve robots is: Flexible picking and handling operations.
Steel robots can only see the size of an object, aiming where each finger is placed to grasp the object, but people do not do so. When we want to take something, reach out and grab it in a very continuous motion without thinking about size or which finger to use. Because of the precise requirements placed on the position of the fingers, robots have a lot of limitations in their ability to pick up, and they have no way to deal with irregular objects.
The soft move came into being. Because there is no need to look closely at where this object is placed, it will not be controlled by the shape. For example, it can grab eggs, paper strips. Because this robot has a very flexible structure, it is free to deal with various uncertainties.
We can also let robots have the ability to distinguish real objects by embedding some simple sensors. Of course, this cannot be done 100%, and the recognition accuracy in some scenes is low. Horizontal or using two fingers will make the success rate higher, because horizontal grabbing accumulates more data and knows how to grasp; and two points grabbing information is less.
Language Exchange
Even if there is a software structure, sometimes the robot will fail. why? If the robot can't catch it, it can tell people what's wrong, but it can't.
Through observation, it can be found that when a robot performs a task, human intervention will completely change its plan. How to improve human-computer collaboration and interaction? If the robot can simply say "help me, I'm stuck," this will solve the problem, but at the moment it can't do it. In addition, if the robot can also introspect, calculate new decision-making actions based on their own data to avoid this failure.
So we want to give robots this ability. We have developed a procedural planning system through which robots can think about their own course of action—when stuck, think about “why do you get stuck and how can you get out of this hurdle†or communicate this idea with humans? "Please move the table."
So imagine that robots must have the ability to communicate and communicate with the outside world very clearly and clearly. Otherwise, it can only say "help me," and humans have to check to see if it has any problems in the end, so that efficiency is very low.
Cloud Big Data helps learning
We know that robots also need to learn. However, we humans can accept a large amount of data every day from birth to learn, and for robots, data storage is easy to lack of memory. A self-driving car has 1TB of data in one hour and it is difficult to analyze. Therefore, we need to increase the degree of abstraction so that the collected data can reach higher levels and reduce the storage pressure and the amount of computation.
For example, on the left is a GPS data stream. If we can establish a meaningful structure for this GPS data stream, we can summarize some extractable information and then do high-level reasoning. For example, when you drive to a certain location, you know what tasks to perform.
Extracting data from the data stream, abstracting it, and summing up meaningful information—this is the core technology to share next—using an algorithm to analyze small datasets in big data. These small datasets The set can reflect the entire data operation result.
The same example: we use Coresete's method to analyze the video to get the number of data sets, then focus the different colors, from which we can analyze more and more complex video. For 16500 frames in the movie, we only need to analyze 1152 Coresete data points.
Multi-robot system
With only one robot, the tasks that can be accomplished are limited. We need many robots to form an automated system. So, the fifth trend is the multi-robot system.
When several robots come together, each robot has its own job. Of course, if you are building a cabin now, one of the robots will be responsible for moving parts, and the other robot will be responsible for other work. So we can see that these four robots are collaborating.
The robots must be able to communicate and coordinate with each other before they can know when to perform tasks with their peers. This is a challenge. They need to understand their own tasks, but also know the situation of the entire collective task.
On-demand manufacturing
Our goal is to have a robot print directly out of a 3D printer, but this is not a general shell print. There must be a drive mechanism in the 3D printer. We can see the electronic structure inside. This is actually a very complicated mechanism.
Popularization
Let everyone design their own robots. Is this idea crazy? With databases, programming tools, 3D printing, and other technologies, although not all robots can be automated, it is true that many steps can be automated.
At the same time, robots have in fact a very wide range of uses that can permeate all aspects of our lives. For example, if we swallow a tiny object by mistake, we can make a micro-folding robot and send it to the intestines so that it can wrap foreign bodies through the folded type and take them out of the body. This helps us to avoid minimally invasive surgery; Or, you can use micro-robots to provide stomach treatment to humans.
Learn robots early
Let students start robot learning early and use programming tools to create a variety of robots. The goal we hope to achieve is to use robot magic to attract students - not only robot shells, but also to learn programming and other soft capabilities. Let students enter the world of robots happily and gradually invest in the robotics industry.
Competitive cooperation between academia and industry
What we are facing now is an unprecedented computer industry revolution. We need to have forward-thinking or whimsical ideas in the academic world. We also need the cooperation of industry to make these ideas into products. At the same time, the government should also participate in it and put forward the correct implementation plan so that the robot can really play its role. The United States DARPA is a good example.
A year ago, Toyota Motor found MIT. The company stated that driving is fraught with danger, and that the number of people who have lost their lives worldwide due to traffic accidents has reached 1.5 million every year. Through industrial and academic cooperation, the wisdom of academics is applied to the difficulties of the industrial sector so that smart driving vehicles can be better developed.
Autopilot
MIT is also conducting research and development of smart-driving cars. In mid-2010, Singapore proposed such a program. The combination of smart driving and automated taxi service provides an automatic car network for the city.
However, in general, we can only drive in a simple environment. There are still many obstacles to overcome in a real driving environment.
Business investment and entrepreneurship
Now the world is also aware of opportunities in the robotics field. In recent years, large-scale investments have taken place in the robotics field. In 2015, there were approximately 2 billion U.S. dollars worth of M&A transactions.
China's innovation
China really needs to lead the revolution in robotics. I hereby put forward a vision. In the future, there will be more robots and workers working side-by-side. In the future, robots will be more advanced than they are now.
In fact, we need not worry about robots replacing us. In fact, we should worry more about the fact that we have not built the robot fast enough. In China, only 20 percent of people are still working at the age of 2050. Therefore, it is even more necessary to speed up the production of robots to make up for the shortage of labor.
The Massachusetts Institute of Technology (MIT) has been a pioneer in the research of robotics technology. This laboratory has studied cheetahs, Atlas and other sensational military robots. Then, with the development of cutting-edge artificial intelligence technologies such as DeepMindAlphaGo and Atlas, what new trends will emerge in the robotics research? At the CCF-GAIR Global Artificial Intelligence and Robotics Summit Robotics Workshop, MIT Robotics Laboratory Director, IEEE, AAAIFellow, and Daniela Rus, Academician of the National Academy of Engineering, made a presentation on this and described the twelve leading technology trends in the world of robotics: