You should know about these "dry goods" for driverless driving.

The term unmanned seems to be the hot search term of 2017 together with the words such as the Internet of Things and artificial intelligence . A few years ago, you might think that driverlessness will only appear in American science fiction movies, especially in the car of the genius genius scientist of Marvel Iron Man.

Yes, you didn't guess wrong, Marvel Iron Man's car: Audi R8, does have an unmanned module, and dare to put these cars for science fiction movies directly on the table, as a technology to flaunt their own Audi, this What black technology has it taken out? First of all, before talking about Audi's super-running R8 e-tron, let's talk about what is unmanned.

Unmanned, a complete system that can be completely and safely run by a computer without the intervention of the driver. The features are simple and safe, and can be used for automatic parking. A driverless car is a smart car that senses the road environment through an in-vehicle sensing system, automatically plans driving routes, and controls the vehicle to reach a predetermined target. It uses the on-board sensor to sense the surrounding environment of the vehicle, and controls the steering and speed of the vehicle based on the road, vehicle position and obstacle information obtained by the perception, so that the vehicle can travel safely and reliably on the road. It is a collection of automatic control, architecture, artificial intelligence, visual computing and many other technologies. It is a product of computer science, pattern recognition and intelligent control technology. It is also an important indicator to measure the scientific strength and industrial level of a country. And the national economy has broad application prospects. Especially in public rail transit, it is possible to arrange the departures and the labor cost more intelligently and scientifically, so that passengers can get a more efficient travel experience.

Speaking easily, the concept of driverless, Google has been in contact since 2009, and now it has been eight years, but the eight years of technology accumulation is still unable to push it to commercial use, see the development of autonomous driving technology and not simple. Autopilot is a huge project that integrates many hardware and software. Let's first talk about the hardware that needs to be used: car, controller, CAN card, GPS+ inertial measurement unit, sensory sensor. Among them, the car as a carrier of the automatic driving system is naturally indispensable. However, the choice of the car is also very particular. Try to choose a hybrid or pure electric vehicle. In addition to the large amount of power required for the automatic driving system, the underlying control algorithm of the engine is much more complicated than the motor. Instead of spending time researching and debugging the underlying layer, it is better to Develop a higher level motor algorithm. The controller, whose purpose is to control the operation of each unit, is also indispensable. The CAN card is the language used for the "communication" between the controller and the chassis of the car. The CAN card can be simply understood as the "modem of the chassis and controller." Simply put, any condition of the chassis needs to be resolved to the controller through the CAN card, so that the controller can adjust the vehicle through the information collected by the sensor. The global positioning system and inertial measurement unit enable the unmanned system to accurately position the vehicle, while the inertial measurement unit provides physical information such as angular acceleration and yaw rate. This information helps the driverless system to make better decisions and control the vehicle. Perceived sensors, I believe that everyone has seen it on some high-end flagship models, including countless cameras, infrared sensors, laser sensors, radars, etc. Its purpose is like the human eye, collecting relevant information, which is conducive to the brain's decision on the current environment.

(The picture shows the "CAN card" of the Audi driverless system)

The software aspect is based on perception, integration, decision-making and control. Each dimension needs to write code to realize information conversion. Because it is too complicated, there is not much to write here.

Understand the basic principles of the automatic driving system, and then talk about the level of automatic driving, yes, automatic driving is also graded. It is divided into six levels. The six-level classification system for autonomous driving was released by the Society of Automotive Engineers (SAE International) in 2014.

(The picture shows the SAE classification standard)

In the SAE classification standard, most of the daily cars belong to the 0th to the 1st level. Collision warning belongs to level 0; automatic anti-collision and cruise control belong to the first level of assisted driving, and the automatic parking function is between level 1 and level 2. The Autopilot assisted driving technology that Tesla is selling belongs to Level 2 technology. Between the second and third levels, there is a considerable span. Between the 0th and 2nd levels, the driver needs to observe the road conditions at all times, but the task of monitoring the road conditions in the third level is provided by the automatic driving system. carry out. This difference is undoubtedly huge. Of course, in the third level, humans are not saying that they can sleep in the car, go online, watch movies, according to SAE requirements, even in the third level, the driver needs and should be ready at any time in response to the system can not make The request for judgment.

Even if there were warnings beforehand, in the experiment done by Google, there will still be drivers who will give control to the system in the second or third stage because the driverless system performs well, and they will do their own thing. Therefore, in the long run, the fourth and fifth will have commercial conditions and broad prospects for development.

When talking about driverlessness, the question people think of is "is it safe?" There is no doubt that Google's Waymo autopilot system has developed relatively mature and is infinitely close to commercial use. But it is not technical issues, but ethical and policy, social and legal considerations.

First of all, in terms of ethics, because human beings stand on their own moral points to look at unmanned driving, that is, if unmanned driving causes death, even if the number is less than the death caused by human beings under the same conditions, it is unacceptable. . Even if the development of unmanned technology is good, there are casualties caused by certain uncontrollable factors (extremely harsh road conditions, specific scenes that the machine has never learned), and sometimes even fatal. The only reasonable solution is to reduce the chance of accidents caused by driverlessness as much as possible. For example, there is only one accident in 100 unmanned driving, which may be better than 5 accidents in 100 human driving. Acceptable.

Secondly, the legal aspect, because the current legal framework is formulated with humans as drivers, does not take into account the disputes arising from driverless driving, and road design, road conditions, insurance policies, etc. are also based on human beings as the sole driver. If you want to make unmanned technology popular, the corresponding laws, regulations and policies must also keep pace.

Finally, the most realistic reason is that the impact on the traditional industry, no need to drive people, it means that some people want to lose their jobs. Let's not talk about the living conditions of the network car and the taxi driver. In terms of trucks and truck drivers, the group of truck drivers is huge in both the United States and China. Nowadays, it is also indispensable. And the circulation of goods depends on the truck driver. This is a huge amount, and it is conceivable that it is difficult to popularize driverless trucks.

As far as China is concerned, compared with other countries in Europe and America, there is a certain advantage in popularizing unmanned driving. First of all, China is in the stage of rapid development of the transportation network. China is easier to start from the perspective of road construction than other countries, equipped with special roads, traffic signs and even targeted traffic regulations for autonomous vehicles. This can make up for many of the shortcomings of autonomous driving technology itself, and the risk of accidents in autonomous driving technology is greatly reduced. In fact, China's resistance to the development of high technology is not as great as that of European and American countries. Concentrating power to do big things is also one of the characteristics of China's development. The fastest and longest high-speed railway in the world can be built in just a few years. For China's current poor traffic conditions, most of them are caused by human factors, all kinds of chaotic parking, driving according to the rules, changing lanes at will. At the same time, drunk driving and fatigue driving have a high mortality rate in China. The application of the automatic driving system is of great significance for improving the traffic situation in China and reducing the number of human traffic accidents. In short, China plays an important role in promoting and popularizing driverlessness.