Image source @ Visual China
Wen | talk about AI, and the author | Zheng driving
Jidu's auto robot is one step closer to us.
Recently, an episode during the exposure of JIDU SIMUCar 2.0, presumably everyone has heard about it, yes, the dual lidar scheme in its detail map fell sharply on the front cover, and also attracted the attention of a big name in the industry.
Ideal Car founder Li Xiang sang two opposites to this design, one is that the lidar roof or two covers, there is no difference in performance, and even the performance of the single roof will be better. Second, the car cover scheme, the latest pedestrian collision regulations, this pass can not pass.
However, in this regard, Jidu CEO Xia Yiping and Jidu also quickly refuted, first of all, Xia Yiping believes that the two FOVs on the cover are greater than the one on the roof, and there will still be differences in product capabilities and safety.
In terms of concentration, it is also said that "the lidar function is very important, but the layout must also consider the beauty of the project."
The in-laws between the big guys have their own reasons, in fact, in the new car to today, there are already some strange things. After all, on many details, everyone is doing the first person to eat crabs.
However, the design of Jidu is indeed rarely seen in the industry, somewhat like the tears of the "jumping lights" era, even if it cannot be accepted by everyone, but the differentiated retro and technological integration of visual sense is undoubtedly a selling point.
After all, just like Rolls-Royce, if the future puts lidar on the iconic celebratory goddess, the goddess who has been static for about a hundred years has begun to release scanning harnesses from her eyes, which is cyberpunk flavor, thinking about it is a bit of a hanging sky.
But just because of the appearance to be the first crab eater, obviously too risky, for the concentration, where the lidar is installed, in the end, it is still a trade-off problem.
Regarding Jidu, another point worthy of our attention today is that according to its product positioning, what they want to build is not a simple car, nor a smart car, but a car robot that sounds quite tall.
So can the still hazy concentration of today's products carry the special connotation behind the word car robot?
About the gap and expectations
Robin Li said at the end of last year that Jidu's auto robot runs through the three major product concepts of "free movement, natural communication, and self-growth", which point to the ability to automatically drive at the L4 level; the accurate recognition of human-vehicle interaction and semantics; and the ability of auto robots to learn and iterate on their own.
Such a product concept is indeed like the existence of open hanging in the current industry, but for Jidu, Robin Li's words are actually more like the outlook for the future of the product.
Because in today's view, the first mass production model wants to fully implement these three product concepts, which is really unlikely.
Take the "free movement" in the first place, the set degree of automatic driving ability is still relatively sufficient, after all, what Baidu Apollo has been doing is to explore the boundary of unmanned driving, and the technology decentralization is somewhat dimensionality-reducing.
But Xia Yiping once admitted, "Baidu Apollo is the highest configuration, applied to the mass production car needs to be cut, mass production is actually between the function, cost and the final realization, to seek a balance." ”
Tan Qing said that AI believes that behind this balance, what is reflected is actually the clamping of multiple factors such as regulations, ethics, and costs at the practical level.
First of all, at the regulatory level, even if the set degree can be L4, there is currently no open right of way for mass production models of L3 and above in China, and when it comes to the market, it is undoubtedly necessary to downgrade the technology.
Or even if the country is open, the ethical problem behind it is enough to drink a pot, because on this issue, he must be the first to eat crabs.
Why is it said that Jidu must be the first to eat crabs? Don't take the licensed Mercedes-Benz L3 seriously?
It is really not needed, Mercedes-Benz L3 certainly puts the responsibility on itself, but if you look closely at the practical rules, you will find that you have been clearly arranged by Mercedes-Benz.
Because if you want to make Mercedes-Benz responsible, you must first take advantage of the time and place.
It's almost only available on sunny days; it's convenient to only open to 8,000-mile highways in Germany; and it's necessary to drive at speeds below 37 miles (60 km/h).
It is not difficult to find that it is indeed a technical job to use automatic driving to hold Mercedes-Benz responsible.
This is actually a microcosm of the industry's ethical issues about responsibility, because to a large extent, the ethical barriers that this technology has to pass will not be completely transformed in a certain moment.
So in the future, he will have to give consumers those "traps" in the division of responsibilities in the pioneering period? This seems to be an unavoidable problem.
Next, if you want to achieve "free movement", the cost is also a challenge that the set degree cannot ignore.
The Mercedes-Benz L3 is currently equipped with only the S-Class and EQS, two million-dollar luxury cars, which are the premise elements of its inclusive expensive technical costs.
Not to mention the L4 such as the L4, in order to pile up all aspects of redundancy, most players such as Waymo, Baidu and so on are currently investing at almost no cost.
Therefore, as Xia Yiping said, if the pricing of Jidu products cannot help it tolerate high costs, then cutting the Apollo L4 is inevitable.
This seems to mean that Jidu can be described as a product concept that satisfies the appetite of consumers, and it is likely that the first mass production car will bury the greater the hope and the greater the disappointment of the hidden worry for the market.
But this problem is actually more inclined to the marketing or public relations level, the final competitiveness of the color, or fall back to the product.
After all, jidu ultimately has to challenge not the outlook, but the more brutal table in reality.
From this point of view, Jidu proposed the concept of automotive robots, in fact, it will not make people too confused.
On the one hand, the greater the expectation, the greater the disappointment, and the concentration is self-aware, and he still insists on doing so, which seems to let us see that he is full of confidence in the product.
In addition to the technical blessing of Apollo, Jidu also has a Baidu map with greater advantages in ToB, Geely's ready-made vast architecture, intelligent chassis technology in cooperation with ZF, etc., which is indeed a "dream start" compared with many new forces that have struggled and traveled a long way.
More importantly, Baidu has been deeply involved in autonomous driving for many years, and As the first crystallization that can show Apollo's strong skills, He really needs to set up a higher market expectation for himself.
Therefore, the set of bullets is worth flying a little longer. However, this bullet should not fly for too long, because he is really, too fast.
The tripartite consensus of "only fast is not broken"
Set him really fast.
Tianyancha APP information shows that on March 2 last year, Jidu was officially established, and four months later, the shape of the first car was confirmed; five months later, the core supply chain system formed by Ningde Times, Continental Group, Hesai Technology, etc. was completed; six months later, SIMUCar road test exposure.
Today, he has been holding the pipa half-covered, and Xiaomi, who debuted to build cars at the same time, is still laying the foundation for the factory.
Regarding such a fast and unbreakable car-making strategy, it seems that a considerable degree of consensus has already been reached between Jidu, Baidu and Geely.
Demand for beach grabbing
A few days ago, Tan Qing said that AI saw that there was a view that the new forces of car manufacturing are just paper tigers, and the century-old car manufacturing heritage of traditional car companies is doomed to make them out of the king.
The reason why such a surprising statement is that the traditional car company products often seek stability, research and development is fully mature to push new, the new forces are first radical listing, and then sewing and patching, the former has a harder foundation.
This view seems to be a bit alarmist, but in fact, it does not stand up to scrutiny, after all, Mercedes-Benz took the lead in introducing the L3, which seems to be less conservative.
Tan Qing said that AI believes that using thick accumulation and thin hair to describe the slowness of traditional car companies is more likely to be a high emotional intelligence answer, but the fact is likely to be a conclusion of low emotional intelligence, that is, they are fighting back in the vortex of passive and helplessness.
If the era of intelligence still blindly pursues "slow work and fine work", then the first person who is dissatisfied may be Boss Jia.
Faraday's future car was really good a few years ago, and if Boss Jia had not "suffocated" that year, he might have ushered in the second spring of his career.
But today is still difficult to give birth, compared to the competitors on the market, today's Faraday's future is already like Faraday's past.
Tan Qing said that AI believes that when we agree with the truth of slow work and fine work, we often have to be based on the fact that we have to practice it without competitors.
Today, although Jidu lies on the hotbed of the double giants, but the technology also has a shelf life, from Tesla to the new forces, to the autonomy, traditional car companies, as well as in the future millet, Sony and other cross-border big brothers, Jidu really needs to race against time.
Tan Qing said that AI believes that this problem is actually like the plot of the last killing of Huang Lang in "Let the Bullets Fly", seeing can be true, the best way for Baidu to mobilize investor confidence at this stage is not to have more technology, but through mass production, let Apollo carry out a flood of strength, only in this way, Baidu's technology that has been deeply cultivated for many years can find a breakthrough that no longer allows investors to keep waiting.
More importantly, high-level automatic driving has entered a bottleneck period at the technical level of bicycle intelligence, which further intensifies Baidu's demand for rapid attack.
Geely's new battlefield
Looking at Geely, its strategy has also been tilted since last year.
Although many autonomous car companies were born in the era of joint venture cars, they have also experienced a vicious battle, and their sense of smell of crisis is usually more acute. However, in the early years of Geely, the burden of electrification transformation still existed objectively.
Different from today's many traditional car companies or acquisitions, or cooperation models to rapidly develop intelligence, Li Shufu in 2016, it seems that there is still some self-research inertia, choose a small fire slow stew.
Its early establishment of the automotive intelligent technology company Yi Jiatong, mainly to control the car chip this fortress, Baidu also led it over 1.3 billion yuan A round of investment, but to this day, the volume is still limited.
Since last year, Geely seems to have discovered the importance of racing against time, not only with Baidu to establish a joint venture, Geely also announced this year that it chose Mobileye, a partner who also does chips and autonomous driving systems, for the Extreme Krypton brand.
Not only that, the automatic driving chip equipped by Jidu today is not from Yijiatong, but NVIDIA.
Tan Qing said that AI believes that from geely's attitude change, we seem to see that the current EGA AND its chip companies have not yet reached the level recognized by Geely in terms of strength.
But more importantly, Geely seems to have made its own trade-offs between control and speed.
From the perspective of the support and let the two big trees behind the set, the best automatic driving technology and electrification architecture, we all give, our own chips can not work, then do not force, use others. Such two big trees do not seem to want Jidu to become a paddling brand in the future.
Therefore, Jidu may really want to go to this dream of "car robot".
We can't deny that for The Set, who has just celebrated its first birthday, it is a commendable thing to be able to develop rapidly to this stage.
But in the face of brutal competition, will he be able to become famous in next year?' The answer is still up to the market.