The Baidu World Congress saw the release of many of the company’s latest research developments. Among this pile of content, the most worthy of commentary, I think, is the latest release of the unmanned car Apollo RT6. The RT6 released this year is very different from the previous Apollo series of cars, the previous Baidu cars are Apollo installed on other people’s cars.
This year’s Apollo RT6 was made by Baidu itself from start to finish.
The biggest selling point of Apollo RT6 is that it is a truly “unmanned car”, buy it without you driving, it will move on its own.
And the field of autonomous driving is Baidu’s strength, the car is equipped with the latest L4 level of Apollo OS, longer training than the previous generation, and better car skills.
Its safety is also a point of concern for many people, for this aspect, the industry has a special term called ” safety redundancy “.
That is, the more technology that can ensure safe autonomous driving of the vehicle, the better.
In Apollo RT6, Baidu has made seven safety redundancies for the car from start to finish, which can be said to be the first in the industry to take care of everything.
After this launch, there were quite a few people who were surprised.
After all, we all know that Baidu’s research on autonomous driving for quite a few years just did not expect it to “fast forward” to the car, and not from the ordinary new energy car to build, directly zoomed in on the launch of the unmanned car.
You may feel that this car and their owner have nothing to do, but maybe a few years later you want to take a taxi out, parked in front of you may be a RT6.
These years, ” self-driving ” which originally belongs to the science fiction film technology from our lives has become closer and closer.
Why do you think we are so eager to self-driving it? Is it just because it looks ” more science fiction “and more advanced?”
Obviously not, the advantages of autonomous driving is actually very simple, the focus on one: efficiency.
I think many drivers know all about the various kinds of traffic jams.
In addition to unexpected car accidents, reckless lane changes, low-speed driving and other reasons will make the original navigation all the way through the green into red.
Unless the road planning is particularly stupid, in fact, today’s urban traffic management system is already very intelligent, if everyone obeys the traffic rules will rarely appear congestion.
The AI responsible for autonomous driving is absolutely “law-abiding” in this regard, the 60 yards will never open 50 yards, and will not speed indiscriminately so that car owners eat tickets, which naturally improves the operational efficiency of the entire traffic system.
The benefits of autonomous driving are so many, it is also a major trend in the future of travel, there are predictions that by 2030, the autonomous driving market may reach trillions of yuan.
In fact, in the industry of autonomous driving, there are already many companies doing things like unmanned trucks, unmanned delivery, and unmanned buses.
Because the level of autonomous driving technology required in these industries is relatively low, there are many companies that can meet the standards.
In contrast, the unmanned cab field is such a big cake, so far no company can take a piece of it, what is the reason for this?
For example, there are Waymo of Google, Cruise, Uber and other companies in the field of autonomous driving reputation, and there are also veteran companies like Baidu that have been involved in autonomous driving for many years.
But whether domestic or foreign companies, in the road to commercialization of self-driving landing, eat the pain is actually similar.
One of the most important or policy regulatory issues.
The policy for the unmanned cab regulation is strict is no problem, after all, human life is at stake, whether it is for the safety of passengers, pedestrians or drivers, this matter can not be sloppy.
But stuck to keep these companies from testing on the road is not a solution, after all, AI training needs data to support it.
The leaders of the relevant departments in each country can naturally see this problem, so they are also slowly liberalizing regulation.
Whether at home or abroad, there have been a number of small-scale pilots, so that these companies can first try the water.
Like Baidu took this opportunity to launch the radish fast run, in Beijing, Chongqing, Wuhan, Yangquan four places to start commercial charge operation.
So far, the number of orders for carriage has exceeded 1 million orders, and there is even a user in Changsha who has played more than 500 times.
And Baidu also became one of the first companies in China to get the unmanned manned demonstration application license in Beijing a while ago, and opened an unmanned self-driving travel service on the same day.
In the future, in the areas specified in the policy, get in the car, the main driver does not need to sit in a person.
If Baidu wants to continue promoting unmanned cabs after this, it will depend on which lucky cities will be the new pilots.
But don’t worry, I heard that Baidu plans to operate in 65 cities by 2025 and will cover 100 cities by 2030.
In addition to policy regulation, there is actually another thing that is also a big headache for self-driving companies, and that is the high cost of self-driving cars.
The first is the software and hardware, long-term AI training, and road test data acquisition how laborious not to mention the effort.
Just a self-driving cab fleet of 50 cars, driving 6 hours a day to generate data to reach 1.6 PB, the ideal commercialization, the number of cars to add three or four zeroes.
This huge amount of data processing is enough to give these companies a headache.
In addition, for self-driving cars, covered with the body of LIDAR, cameras, ultrasonic sensors and other ” eyes ” is the most important organ, they determine the judgment and accuracy of self-driving cars.
These things must not be second best, and therefore are quite expensive.
In addition, the vast majority of unmanned cars want to commercialization are dependent on third-party diversion such as taxi platforms, which is another large cost, not to mention the cost of the vehicle itself and other additional costs.
So a round of calculations down, where to spend money, but Robin Li’s expectations for unmanned cab travel is cheaper than the current taxi service, the cost of how to press down?
These issues have obviously been considered by Baidu, which has many advantages over others in these areas.
In terms of policy, Baidu has a wealth of experience in autonomous driving technology, and the first batch of top-level road test licenses issued in 2019 are all for Baidu to take.
This makes Baidu a leading road test data, and the Chinese road environment is more complex and variable, the training effect is also much better than in foreign countries.
And Baidu’s ability to process data is also very strong, do not forget that the old business is to do search engine origin.
As for the cost of the vehicle, Baidu Apollo RT6 data simply blinded me: the cost of 250,000 yuan, almost a tenth of the peers, this is how to do it?
After a closer look, I realized the doorway here.
Once most of the car first care is ” driving experience “, so there are many luxury features designed specifically for the driver.
But now this is an unmanned car, since no one has to drive, so the money of these luxury features can naturally be saved, and put into the rear passenger comfort improvements.
Elsewhere, Baidu is also controlling costs through various methods.
There are three main aspects, first, the whole vehicle cost reduction. Customized for operational scenarios, Baidu Cloud applies a fully-owned intellectual property platform to significantly reduce the cost of the vehicle itself.
Second, the cost of the autonomous driving kit is reduced. Based on technological advances, the RT6 uses vehicle-gauge mass-production sensors and computing units, significantly reducing the cost of autonomous driving kits.
Third, the advantages of manufacturing in China, supply chain, and manufacturing maturity and perfection, further reduce the cost.
Not to mention that Baidu also launched the turnip fast running platform, the first batch of Apollo RT6 is also placed on this for passengers to call, there is no middleman to earn the price difference, naturally can reduce operating costs.
So look down, other self-driving companies in the future commercialization of landing may encounter problems, Baidu has basically thought ahead and solved, perhaps China can really become the first country to achieve the “future of travel”.
I remember watching a movie called “The Dragon” when I was a kid, and there was a scene where the female lead was talking to the self-driving car, which made a deep impression on me.
I used to think that if I could really meet a self-driving car in my lifetime, I would have to talk to it.
Now I’m still looking forward to self-driving cars, but I don’t think I’ll be talking to them anymore. I think as long as they can let me sleep for another half hour on my way to work, that future is enough to make me smile.
For people living in big cities, if unmanned cabs are really popular, they will no longer have to worry about being late for work for a few minutes.
In the future when unmanned cabs are popular, even if you are not a commuter, you can really feel the convenience it brings, because unmanned cabs have cheaper prices, safer and more standardized taxi service, and a more comfortable environment.
Such a future, I hope it can come soon.
