标签归档 南京夜生活

The scorching summer has come, and other creatures on the earth are also suffering from high temperature besides human beings. You may not realize that the ingredients that are about to become delicious food in your mouth are also in danger of "heatstroke". For example, refreshing coffee, cool orange juice and sweet chocolate. The book "The Earth Doesn’t Care: The Table Destroyed by Climate Change" shows readers an unknown global food picture, and those foods that we are used to may become increasingly scarce in the future.　

The earth doesn’t care: the dining table destroyed by climate change

Coffee that continues to reduce production.

Since 2010, the global coffee bean production has been declining, because the main coffee producing areas are located in Africa and Brazil, where the temperature is rising, especially the temperature difference between day and night is getting smaller and smaller, which directly affects the growth of coffee trees, and coffee gardens have to move to higher altitudes.

Some studies have predicted that Arabica coffee, the queen of coffee today, will be severely hit in the future. The optimal growth temperature of this coffee bean is between 18 and 24 degrees Celsius. If the temperature exceeds 30 degrees Celsius for a long time, its growth will stagnate.

Arabica coffee

In addition, there are more and more extreme climates. The high temperature scorches the coffee leaves, and the rising air humidity will cause more pests and diseases, such as a terrible fungus, Camellia sinensis. These effects not only reduce the yield of coffee beans, but also increase the price of coffee.

The book points out that if you want to continue drinking coffee, you need to change the coffee planting system. For example, choosing a coffee tree species that is more adaptable to changes requires farmers to pull out the old varieties of coffee trees. Some professors joked that "it is not so simple as separating a man from his wife".

For another example, coffee farmers should learn more sustainable planting methods, including improving soil, mixed planting, setting up cooperatives and so on. However, the change of agricultural production can not be achieved overnight, and it needs long-term experience accumulation, but also needs various considerations of social economy. I am afraid that the progress of human-made change can not keep up with the speed of environmental change.

Therefore, the author still pessimistically points out that in the future, only those who can still afford high-priced coffee can casually promise "Let’s have a cup of coffee!" For ordinary people, one day in the future, they may have to give up the leisure hobby of coffee.

Oranges plagued by plague

80% of the concentrated orange juice in the world comes from Brazil. However, a plague, Huanglongbing, broke out in orange plantations in Brazil. For oranges, Huanglongbing is just like Covid-19’s threat to human beings. The plague has swept the world, and it is constantly mutating in the process of spreading. According to FAO data, Huanglongbing has destroyed more than 100 million orange trees worldwide.

Citrus trees with Huanglongbing disease

The pathogen of Huanglongbing first lived in the hot environment of Asia, and was first discovered by farmers in Chaozhou, China hundreds of years ago. In 1965, the pathogen was found in southern Africa. Two years later, experts found its trace in the Philippines, and the pathogen and citrus psyllid combined to form a more harmful disease. In 2004, the disease also appeared in Brazil, and a new bacterial variety was produced, which began to sweep the whole America.

This plague will spread from orange tree to orange tree. When they are infected, they will give off the fragrance of methyl salicylate, attract a large number of psyllids, and further spread the disease. The researchers pointed out that with the increase of temperature, the reproduction speed of psyllid can be increased from 4 generations to 8 generations per year.

Huanglongbing not only leads to the decline of orange yield, but also makes the taste of fruit sour. In other words, orange juice is not as good as it used to be, and the composition of juice must be redesigned. This has brought new challenges to the orange juice industry.

Like human beings, solving the plague problem requires a high cost. In fact, except oranges, many crops are facing the same disease disaster problem. It is also mentioned in the book that vineyards in Europe are being harmed by a fruit fly, and olive trees are also being harmed by another fruit fly.

More and more "luxury" chocolate

Cocoa trees generally grow between 20 degrees north latitude and 20 degrees south latitude. Cocoa beans produced in C? te d ‘Ivoire and Ghana in western Africa meet the demand of two-thirds of the world cocoa bean market. With the climate change, the local rainfall is getting less and less, and cocoa trees can’t stand the long-term high temperature and dry climate. The optimum growth temperature of cocoa trees is 21-32 degrees Celsius, which can withstand the high temperature of 36 degrees Celsius for a short time. Once the temperature is exceeded for too long, the cocoa trees will suffer from "heatstroke". If there is not enough water to cool down, their own circulation system will soon collapse.

cacao tree

According to the research report "Vulnerability of Cocoa Trees in West Africa to Climate Change", the whole West Africa cocoa will be brought to the middle of the 21st century, and 50% of the planting area will disappear due to high temperature and heat. In 2016, The Wall Street Journal speculated that the chocolate industry had invested about $1 billion to develop measures to deal with climate change.

Edward George, an agricultural expert, said: "Maybe one day, chocolate will really become a luxury like champagne in Europe, and it is no longer something that ordinary people like us can easily imagine."

In fact, like coffee trees, orange trees and cocoa trees, most agricultural products planted by human beings are facing the same dilemma, with less and less water resources, more and more harsh living conditions for plants and more and more pests and diseases. At the same time, the global population is still growing, and the demand for food is increasing. In the future when climate change is getting worse, it may become more and more difficult for human beings to fill their stomachs.

Casino dice

In order to solve the food problem, we must have a deep understanding of the most serious problem facing mankind-climate change. However, in the view of William .Nordhaus, the winner of the Nobel Prize in Economics in 2018, the agricultural problems concerned in The Earth Doesn’t Care: The Table Destroyed by Climate Change are still controlled systems under the influence of climate change, which can be solved by scientific and technological means. That is to say, it is still possible to make crops grow in a completely artificial environment, and of course the cost will be high.

Global warming, storm sea level rise.

Nordhaus classified the impacts of climate change into two categories, and the uncontrolled system is more terrible than the controlled system. Uncontrolled systems include sea level rise, ocean acidification, the intensification of hurricanes and the loss of ecosystems, and their operation is basically free from human intervention. Take ocean acidification as an example. Although some people still question global warming, there is basically no debate about ocean acidification, because it can be verified by simply measuring pH value. Ocean acidification will seriously affect fisheries, and oysters, corals, plankton and shellfish will be affected. Scientists’ research on this topic first appeared in 2004, and it is still in the initial research stage, so no solution has been put forward.

In the public’s cognition, climate change is still too far away. Although extreme weather such as high temperature, drought and rainstorm are becoming more and more frequent, the public still can’t directly understand the impact of climate change on themselves. Some people even think that natural scientists pay too much attention to environmental protection and fail to consider the needs of economic and social development.

Climate Casino: Risk, Uncertainty and Economics of Global Warming

If you feel the same way, nordhaus’s "Climate Casino: Risk, Uncertainty and Economics of Global Warming" will give you a new perspective to understand climate change. As the author’s metaphor in the title of this book shows, human beings are entering the "climate casino" and throwing some climate dice, some of which may be dangerous. So, what can gamblers do in casinos? First of all, we need to know the consequences of different dice to help us get out of the casino safely.

How much will it cost to mitigate climate change?

This popular science book introduces the economic policies to deal with climate change. The most important thing is that the author always discusses the possible economic impact of climate change from the perspective of "cost and benefit", so that readers can understand that climate change is actually an economic problem.

Nordhaus believes that there are three main ways to deal with the threat of climate change.

The first method is adaptation, which means that we should learn to cope with the warming world, not stop it. This method is highly praised by people who think that climate change is negligible. But the problem is that the economic cost of adaptation is not low. If the investment in agricultural facilities is acceptable, then the investment in dealing with sea level rise may be astronomical. Therefore, the general view of experts at present is that adaptation is local and prevention is global.

The second method is geo-engineering, which is generally divided into two categories. The first category is the technology of removing carbon dioxide from the atmosphere, and the second category is the solar radiation management technology that reflects sunlight and heat back into space. But the problem is that geoengineering can not completely offset the greenhouse effect on the technical level; On the implementation level, geo-engineering is a double-edged sword, which has a certain constructive role, but at the same time it is potentially destructive. It can be used in good faith or for malicious destruction. Therefore, the author believes that geo-engineering is similar to what doctors call "rescue therapy" and can only be used when other schemes fail.

The third method is to slow down (prevent) climate change, that is, to take action to reduce greenhouse gas emissions and reduce its atmospheric concentration. This is the focus of current global consultations, and it is the safest solution from the environmental point of view. The author thinks that although it is the most expensive to slow down climate change in the short term, it is actually the most economical way in the long term.

Reduce greenhouse gas emissions

Through the calculation of economic model, nordhaus found that if we live in an ideal world-all countries in this world coordinate and jointly launch emission reduction measures, earnestly ensure that all countries and departments participate, and act in a timely and efficient manner. Then, it is feasible to achieve the goal of "2-degree temperature rise ceiling" proposed by Copenhagen agreement, and it only needs to spend 1%-2% of the world’s total income every year.

However, such a world does not exist. Because "reliable and correct, it may not be popular and realized, and even in reality, it will be fiercely resisted."

Through the economic model, nordhaus found that if only half of the countries participated in the efforts, it would not only be much more expensive, but also impossible. In addition, inefficient control or methods will also double the cost. Therefore, mankind still needs to strive to develop more efficient technologies, design social mechanisms to promote economic efficiency and high participation rate, and provide help to poor countries with limited resources. Finally, we may need to adjust our goals so that they can be achieved.

Obviously, the goal of the upper limit of 1.5 degrees has become increasingly difficult to achieve. On May 17th this year, the World Meteorological Organization (WMO) released the latest report, saying that the global temperature may rise to a record high in the next five years, and it is very likely that it will briefly break through the threshold of 1.5 degrees Celsius, one of the key indicators of global warming stipulated in the Paris Agreement.

How much do you have to pay for carbon emissions?

See here, you should probably understand why many countries have put forward the goal of carbon neutrality, and China also put forward the goal of "carbon neutrality in peak carbon dioxide emissions in 2030 and in 2060" in 2021. You may also understand that what so many international climate conferences are discussing is actually urging all countries to make commitments and take practical actions. Because this is the only choice for mankind at present.

So, for ordinary people, what do you have to pay for it?

On the economic level, nordhaus proposed two ways, one is carbon tax, and the other is carbon trading. Simply put, it is to set a price for carbon emissions, and then let enterprises or consumers pay for it in different ways. In fact, this has already appeared in our daily life, such as the familiar fuel tax. He pointed out that even the most conservative economists support carbon tax as the most efficient way to slow down global warming.

However, in the policies of different countries, nordhaus believes that many control policies are inefficient or ineffective, with little impact, but they need to spend huge costs. For example, subsidies for green technologies provide financial incentives for solar energy and hybrid vehicles. Nordhaus raised his own question, why should we subsidize hybrid cars instead of bicycles? Subsidies are not the solution. It is much more practical to punish carbon emissions than subsidies.

Another thing to note is "energy-cost shortsightedness", that is, we may buy cars with lower initial cost in order to save money now, and underestimate the energy saved in the future. Failure to make the most economical decision is a common feature of human behavior, and human shortsightedness can only see immediate interests. To put it mildly, this is why some low-carbon behaviors are difficult to promote; To put it bluntly, this is the reason why human beings have come to this situation today.

Don’t be a slave to money.

Looking at climate change from an economic perspective will make the analysis and discussion more rational, and of course it will ignore other important contents. After all, money can’t solve all problems. Capitalism in Lishan: Don’t be a slave of money, be a master of Lishan with peace of mind, which is a collection of cases that can broaden one’s horizons and provide innovative solutions.

Capitalism in Lishan: Don’t Be a Slave of Money, Be a Secure Master of Lishan.

This book is based on a series of investigation reports produced by NHK after the 311 earthquake in Japan. They found some alternative economic models and coined the term "Rishan Capitalism" together with the economist Hirosuke Noguchi.

They believe that "money capitalism" is built on the premise that the circulation of funds determines everything, while "Lishan capitalism" quietly establishes a subsystem that does not depend on money next to "money capitalism", allowing people to continuously obtain water, food and fuel when they have no money. Because what people need to survive is water, food and fuel, not just money.

"Lishan capitalism" is a reflection on natural disasters that human beings can’t cope with, and it is also a reflection on the global economic system. The place was originally only a part of the global economic industrial chain, and even a party eliminated by fierce competition. If we can get rid of the inertia thinking mode of "money capitalism", re-examine the local resources and advantages, and even look at the existing problems with reverse thinking, we may find new treasures. The book introduces some interesting cases of "turning waste into treasure".

Energy revolution town

Zhenting City, Okayama Prefecture has carried out the most advanced energy revolution in Japan. The local economy is supported by forestry and wood processing industry. At the end of 1997, the local wood processing factory began to use sawdust to generate electricity, which was called "woody biomass power generation". Sawdust was originally the production waste of the factory, and it needs to be treated as industrial waste, which costs a lot of money. However, since the start of power generation, the power consumption of the factory has been almost 100% dependent on biomass power generation, which has also saved the power consumption cost, which is a typical innovation of turning waste into treasure.

Because of the excessive sawdust waste, the local factory thought of a new use, pressing it into "wood particles" and becoming a new fuel, which can be used by local families and agricultural boilers. Since then, this small town has begun to build a biomass industry and ushered in economic recovery.

In 2022, the European Parliament continued to recognize "original woody biomass" as renewable energy. In fact, Austria is the most advanced country in "power generation from woody biomass". The author of this book also went to Austria for an interview and found that, like Zhenting, a small town in Japan, energy conversion not only solved the energy problem, improved the energy self-sufficiency rate, but also promoted the local economic regeneration and even enhanced national security.

Overcome the lack of society

Aging and declining birthrate are making Japan a "society without opportunities", that is, a society in which people lose contact. There are many vacant houses and elderly people in Zhuangyuan City, Hiroshima Prefecture. The local "Overpopulation Inversion Association" is conducting some welfare experiments on how to turn the problem into resources. They believe that "the disadvantage is not a defect but a treasure chest". Community currency is an interesting attempt.

The staff of the "Overpopulation Inversion Association" heard when chatting with some day-care elderly people that some elderly people like to grow vegetables at home, and they can be self-sufficient, and there will be many vegetables left rotting in the ground, which makes the elderly feel very distressed. So they thought of a way. All the ingredients in the local welfare homes are purchased from foreign real estate. If you buy the dishes of the elderly in the form of community currency, it will not reduce food waste. In the end, the welfare home bought the dishes for the elderly at half the market price, and the elderly could pay for day care with community currency, achieving a win-win situation.

Later, a local restaurant that can use community currency was opened, and a nursery park was opened next to it. The ingredients in the restaurant come from the vegetable garden of the local elderly, and the staff is the mother who is raising children. After sending the children, they can start working. Old people can have lunch in the restaurant or play with the children in the nursery.

Community currency returns to the original function of currency-exchange value, and at the same time undertakes the function of rebuilding local interpersonal connection. People can exchange and communicate with resources and skills that cannot enter the market, especially those who are usually considered as the weak in society, including the elderly, the disabled and housewives.

summarize

In the economic model of Nobel Prize economist nordhaus, we understand why we should reduce carbon emissions to deal with climate change, because it is the only rational choice for human beings. Only along this road, those foods we love can last longer. However, this road is full of thorns, because many problems of "money capitalism" have accumulated. At this time, we need to change some ways of thinking, just like the solution of "Lishan capitalism".

"The insurance premium of Lishan capitalism is not money, but to act and prepare for the matter itself."

For ordinary people, climate change is a problem that we can’t solve with money, but we can use "action and preparation" to win more insurance for ourselves. Preparation comes from our understanding of climate change; Action comes from the change of our daily behavior.

People share the food on the table.

In a recent sustainable food workshop, I shared some information about endangered foods with the participants. I just told them which foods might disappear, and I didn’t have time to introduce the story "The Earth doesn’t care: the table destroyed by climate change" in detail. Even so, when people know that coffee and chocolate may become increasingly scarce, they have shown their frustration that happiness will be deprived.

Perhaps such depression is the seed of "action and preparation". In the face of an uncertain future, we cannot underestimate the impact of climate change, nor can we underestimate the potential of individual action.

Recently, Ashok Elluswamy, director of Tesla autopilot software, gave a speech at CVPR 2022 conference, introducing many achievements made by Tesla autopilot team in the past year, especially the neural network model named Occupancy Networks (hereinafter referred to as occupation network).

He said that there are many problems in the traditional methods of semantic segmentation and depth information used in autonomous driving systems, such as the difficulty of converting 2D to 3D and the inaccurate estimation of depth information.

After using the occupancy network, the model can predict the space occupied by objects around the vehicle (including the space generated by the next action of dynamic objects).

Based on this, vehicles can make evasive actions without identifying the specific obstacles-Ashok Elluswamy even joked on Twitter that Tesla’s car can even avoid UFOs!

Based on this technology, vehicles can also see whether there are obstacles in the surrounding corners, so as to realize unprotected steering like human drivers!

In short, the occupation of the network has significantly enhanced Tesla’s autonomous driving ability (L2).

It is said that Tesla’s autopilot system can prevent 40 car accidents caused by driver errors every day!

In addition, Ashok Elluswamy also highlighted the efforts of Tesla Autopilot system in preventing drivers from misoperation.

By sensing the external environment and the driver’s operating system, the vehicle can identify the driver’s misoperation, such as stepping on the accelerator pedal at the wrong time, the vehicle will stop accelerating and brake automatically!

▲ Tesla active braking

That is to say, some problems of "brake failure" caused by driver’s misoperation that have been frequently exposed in China will be technically restricted.

I have to say that Tesla is really good at promoting technological progress. The following is a video compilation of Ashok Elluswamy’s speech, with a slight deletion.

First, the pure vision algorithm is powerful, and the two-dimensional image is transformed into three-dimensional.

At the beginning of the speech, Ashok said that not everyone knows the specific functions of Tesla’s autopilot system, so he briefly introduced it.

▲Ashok

According to him, Tesla’s autopilot system can help vehicles achieve lane keeping, vehicle following, deceleration and cornering, etc. In addition to these, Tesla’s autopilot system is also equipped with standard safety functions, such as emergency braking and obstacle avoidance, which can avoid a variety of collisions.

In addition, since 2019, about 1 million Tesla vehicles can use more advanced navigation on the expressway, check the information of adjacent lanes to perform lane changes and identify the entrances and exits of the expressway.

Moreover, the Tesla autopilot system can automatically park in the parking lot, identify traffic lights and road signs, and make a right turn to bypass obstacles such as cars. At present, these functions have been verified by hundreds of thousands of Tesla owners.

In the speech, Ashok also took out a video recorded by a user. The video shows that when the user is driving on a crowded road in San Francisco, the car screen shows the surrounding environment, such as road boundaries, lane lines, the position and speed of nearby vehicles.

▲ The system recognizes the surrounding environment.

On the one hand, these need the support of hardware such as Tesla car and camera, on the other hand, they also need the support of algorithms and neural networks built into Tesla autopilot system.

According to Ashok, Tesla is equipped with eight 1.2-megapixel cameras, which can capture 360-degree images of the surrounding environment and generate an average of 36 images per second. Then Tesla’s car will process this information, and it can perform 144 trillion operations per second (TeraOPs/s).

Moreover, these processes are based on pure vision algorithm, without using lidar and ultrasonic radar, and without high-definition map.

So how does Tesla’s autopilot system identify general obstacles?

Ashok said that when encountering general obstacles, the system will use the space division method. When using the space segmentation method, the system marks every pixel in the space as "drivable" and "non-drivable", and then automatically drives the chip to deal with the scene. However, there are some problems with this method.

▲ Marking of objects

First of all, the object pixels marked by the system are in two-dimensional space, and in order to navigate the car in three-dimensional space, the object pixels need to be converted into the corresponding predicted values in three-dimensional space, so that Tesla’s system can establish an interactive physical model and handle the navigation task smoothly.

▲ Marking of objects

However, when the system converts the pixels of an object from a two-dimensional image to a three-dimensional image, it needs to carry out image semantic segmentation (that is, identify the image at the pixel level, that is, label the object category to which each pixel in the image belongs).

This process will produce unnecessary images or unnecessary pixels in the system, and a few pixels on the ground plane of the image can have a great influence, which directly determines how to convert this two-dimensional image into a three-dimensional image. Therefore, Tesla does not want to produce such influential pixels during planning.

In addition, different obstacles need to be judged by different methods.

Generally speaking, the depth value of the object (the distance from the observer’s point of view to the object, which is finally obtained after projection transformation, standardization of equipment coordinates and scaling translation) is commonly used.

In some scenarios, the system can predict obstacles first. In another scene, the system can also detect the depth of the pixels of the image, so each pixel will produce some depth values.

▲ Depth map (right)

However, although the depth map finally generated is very beautiful, only three points are needed when the depth map is used for prediction.

Moreover, when visualizing these three points, although it is ok to look at them at close range, they will also be deformed with the increase of distance, and it is difficult to continue to use these images in the next stage.

For example, walls may be deformed and bent. The objects near the ground level are also determined by fewer points, which makes the system unable to correctly judge obstacles in planning.

Because these depth maps are converted from plane images collected by multiple cameras, it is difficult to produce the same obstacle in the end, and it is also difficult for the system to predict the boundary of the obstacle.

Therefore, Tesla proposed an occupation network scheme to solve this problem.

Second, calculate the space occupancy rate and code the object.

During the speech, Ashok also showed this network occupation scheme by video. He said that as can be seen from the video, in this scheme, the system processes the images captured by eight cameras, then calculates the space occupancy rate of the object, and finally generates a schematic diagram.

▲ The generated simulated image

Moreover, every time the Tesla car moves, the system network will recalculate the space occupancy rate of the surrounding objects. In addition, the system network will not only calculate the space occupancy of some static objects, such as trees and walls, but also calculate the space occupancy of dynamic objects, including driving cars.

After that, the network outputs the image into a three-dimensional image, and can also predict the occluded object, so even if the car only uploads part of the outline of the object, the user can distinguish the object clearly.

In addition, although the resolution of the images captured by the system is different because of the distance, based on the above scheme, the resolution of the simulated 3D images finally presented is the same.

▲ The generated images have the same resolution.

This means that the whole scheme runs very efficiently. Ashok said that the computing platform runs for 10 milliseconds, and the system network can run at the rate of 100 Hz, which is even faster than the speed of many cameras recording images.

So, how is this process completed? This requires an understanding of the architecture of the occupation network scheme.

When explaining the architecture of the occupation network scheme, Ashok takes the images taken by Tesla fisheye camera and the left camera as an example, and compares the image correction process between them.

First, the system will stretch the image, then extract the image features, find out whether the points related to the 3D image are occupied, then use the 3D position coding, and then map it to a fixed position, after which the information will be collected in the subsequent calculation.

▲ Conduct preliminary processing on the image.

After that, the system will embed the position of the image space, continue to process the image stream through 3D query, and finally generate 3D occupation features. Because high-dimensional occupation features are generated, it is difficult to perform this operation at every point in the space. Therefore, the system will generate these high-dimensional features in lower dimensions, such as using typical up-sampling technology to generate high-dimensional space occupancy.

▲ Calculate the space occupancy of objects.

Interestingly, Ashok revealed in his speech that this occupation network scheme was only used to deal with static objects, but in the end it was found that it was difficult to deal with only static trees, and the system also encountered many difficulties when it first began to distinguish between "true and false pedestrians".

But the team finally found that no matter whether these obstacles are moving or static, the system only needs to be able to avoid them in the end.

▲ True and false pedestrians

Therefore, the occupation network scheme no longer distinguishes between dynamic obstacles and static obstacles, but uses other classifications to deal with them and calculate the instantaneous space occupancy rate of objects, but this is not enough to ensure that Tesla cars can drive safely.

Because if only the instantaneous space occupancy rate is calculated, it is not reasonable for Tesla to meet a car while driving at high speed and then start to slow down. The system wants to know more about the space occupancy rate of this car at different times and the changes.

In this way, the system can predict when the car will leave. Therefore, the scheme also involves predicting the occupied flow.

▲ Calculation process of occupied flow

The data of occupancy flow can be the first and higher derivative of space occupancy or time, and it can also provide more accurate control and unify them in the same coordinate system. The system will use the same method to generate space occupancy and occupancy flow, which will also provide strong protection against various obstacles.

Third, the type of obstacle is not important. The system can avoid collision.

Ashok also said that conventional motion or mobile networks can’t judge the type of object, such as whether it is a static object or a moving vehicle.

However, from the control level, the type of objects is actually not important, and the network occupation scheme provides good protection to prevent the network from classification difficulties.

Because no matter what the obstacle is, the system will think that this part of the space is occupied and move at a certain speed. Some special types of vehicles may have strange protrusions, which are difficult to model with traditional technology. The system will use cubes or other polygons to represent moving objects.

In this way, the object can be squeezed at will, and this occupying method does not need complex grid topology modeling.

When the vehicle is turning unprotected or protected, geometric information can be used to infer the occlusion situation. Geometric information should not only infer the information recognized by the vehicle camera, but also infer the unrecognized information.

For example, when a car is making an unprotected turn, there is a fork in the road ahead, and there may be potential vehicles blocked by trees and road signs, so the car "knows" that it can’t see the vehicles from these obstructions. Based on different control strategies, cars can ask questions and eliminate this occlusion.

Therefore, for a stationary object, the car can recognize when it becomes visible while driving. Because there is a complete three-dimensional obstacle, the car can also predict what distance it will hit this object, and then the system will identify and pass this blocked object through smooth control.

Therefore, the occupy network scheme helps to improve the control stack in many different ways. This scheme is an extension of the nerve radiation field, and the nerve radiation field (NeRf) has largely taken over the research of computer vision in the past few years.

▲ Schematic diagram of the association between Nerf and occupied network

NeRf is an image reconstruction scene in a single scene or a single location, which is reconstructed from a point in a single location.

Ashok said that when Tesla’s vehicles are driving, the images received in the background processing are more accurate, so it is possible to generate cross-time and accurate image routes, and generate more accurate 3D reconstruction through the NeRf model and 3D state differential rendering images.

There is a problem with the images in the real world-we will see a lot of unreal or different scenes in the real world.

For example, the glare of the sun or the dirt or dust on the windshield will change due to the diffraction of light, or raindrops will further distort the propagation of light and eventually produce artifacts.

The way to improve the robustness is to use higher-level descriptors, but these descriptors will not change the local lighting artifacts (such as glare) to a certain extent.

Because RGB (color system) images may be very noisy, adding descriptors on RGB can provide a layer of semantic protection to prevent the change of RGB values. Therefore, Tesla’s goal is to use this method to occupy the network solution.

▲ Descriptors are more robust than RGB.

Because the network occupation scheme needs to generate space occupancy in several shots, it cannot run a complete neural optimization in the car, but the neural optimization can be reduced to run in the background, ensuring that the space occupancy generated by it can explain the images of all sensors received by the car when it is running.

In addition, descriptors can also be superimposed during training to produce good supervision for these networks; At the same time, different sensor data can be differentiated to monitor the images held.

At present, Tesla has a network to reduce obstacles. The next step is to avoid any collision. Autopilot already has many safety functions.

Then, Ashok showed three videos of Autopilot starting to avoid collision.

The collision accident here refers to the collision accident caused by the driver accidentally stepping on the accelerator pedal as a brake pedal.

Ashok means that when the driver accidentally uses the accelerator as a brake, the car will accelerate and cause a collision, but the car will recognize and automatically stop the acceleration and automatically brake to prevent the collision.

In the first video, Ashock said that if Autopilot didn’t start and stopped the car from accelerating, the driver in the video would probably fall into the river.

▲ Tesla AP starts to avoid the car falling into the river.

Similarly, the second video shows that a Tesla driver mistakenly stepped on the accelerator while parking, but Autopilot quickly started and prevented the car from hitting shops and pedestrians.

▲ Tesla AP starts to avoid the car hitting the store.

Fourth, automatically plan the path through the occupied vehicle.

However, it may take several seconds or even minutes for the car to brake and stop smoothly, and there may not be enough time to identify obstacles and calculate when the car is driving.

So we should use neural network to achieve this goal; Especially recently, there have been more complicated hidden scenes. What Tesla’s autonomous driving team has to do is to get the space occupancy rate from the previous network.

First, the space occupancy rate should be coded into a super-compressed MLP. In essence, this MLP is an implicit representation of whether collision can be avoided in any specific query state, and this collision avoidance method provides some guarantees in a certain time range. For example, collisions can be avoided in 2 seconds or 4 seconds or within a certain time range.

Ashok gives another example here. He gives a top-down road. Black pixels are obstacles, gray pixels are roads and white pixels are road lanes. In the top view of this three-dimensional space, the car can be placed at any pixel position to simulate whether the collision can be avoided.

▲ Schematic diagram of vehicle driving situation

He said: "If you think of the car as a single point and the period of avoiding collision is set to an instant, then whether there will be a collision at the current time depends only on the position of the obstacle; But the problem is that the car is not a point, it has a rectangular shape and can also turn. "

Therefore, only when the shape is convolved with the obstacle can we immediately know whether the car is in a collision state.

As the car turns (or rotates out of control), the collision field will change. Green means that the car is in a safe position without collision, and red means collision, so when the car rotates, there are more collision positions; But when the car position is aligned, the green position expands, which means that the car will not collide.

On the whole, Ashok shows how to use multiple camera videos and products to generate dense space occupation rate and occupation flow. Through space occupation rate, an effective collision avoidance field can be generated through neural network, that is, vehicles can "see" through the camera and, according to experience, pass through the road with appropriate speed and direction.

▲ Implicit neural network for collision avoidance.

Ashok also shared an experiment in a simulated environment. The driver stepped on the accelerator to accelerate and there was no steering behavior. The car monitored that there would be a collision and planned a path to make the car pass safely.

Ashok said at the end of his speech that if they can successfully implement all the above technologies, they can produce a car that will never crash.

Obviously, this work has not been completed. In his last PPT, Ashok actively invited engineers to join Tesla and build a car that will never crash!

▲ Ashok Elluswamy welcomes more talents to join Tesla.

Conclusion: Tesla continues to explore autonomous driving.

Since Tesla’s self-driving technology with fire, a large number of followers have emerged on the self-driving track. But it must be said that Tesla is still at the forefront of the industry and constantly explores new methods of autonomous driving.

The person in charge of the Tesla Autopilot project brought new technical interpretations, and also showed us the highlights of Tesla’s future autonomous driving technology in advance to some extent. With Tesla’s spirit of continuous exploration, its autonomous driving will continue to lead the entire automobile market.