This article is from the WeChat public account: Geek Car (ID: GeekCar ), author: Yang Jianwen Liu when laughing, title figure: Volkswagen Group CEO Herbert Diess

About automotive software, Volkswagen Group CEO Herbert Dessie once said something like this:

In the long run, I think we might have a bit of an advantage because of scale. On the hardware side, there is probably not so big a difference because they (Tesla) are are also have a dedicated electric platform and they ‘re quite big already for an EV manufacturer.

But when it comes to the next big thing, which is software, Tesla is strong in software – but software really is a volume game. If you do software, you have to use ten million devices, not one million.

Simply speaking, as the head of a car company giant, he first complimented Tesla, and then turned around, calling the car software “the next big thing” and thinking that it was a fight for scale Something that works, Subtext:Tesla doesn’t necessarily have an advantage over Volkswagen.

Who gave him the courage to say this? The answer is the huge human and financial investment of the Volkswagen Group in the software level. Everyone must have heard that news: Volkswagen Group spent huge sums of money to set up the software department “Car.Software”, and at the same time made great efforts to attract software development talents. The goal is to gather more than 5,000 digital experts by 2025 to take full charge of Volkswagen’s automotive software business. In addition, a self-developed vehicle system named “vw.os” will be installed on all models of the Volkswagen Group in the future.

The first is the Volkswagen ID.3, the first production car based on the MEB pure electric platform.

ID.3 exposes software defects, affecting 30,000 units

But reality hits the public.

Recently, the German media reported that there was a software defect in ID.3, and the “number of cases involved” reached nearly 30,000.

We do n’t know exactly what the software problem is, but the solution mentioned in the news: the first batch of 10,000 ID.3 produced will be placed in the parking lot, and the car software will be carried out by the public early next year. Upgrade manually. In other words, these cars do not have the ability to solve software problems through OTA. The next 20,000 cars will be updated by OTA.

These ID.3 will not officially begin shipping to consumers until the middle of next year.

Seeing this news, we thought of Weilai ES8. Weilai also developed its own software system like Volkswagen. During the early delivery, the ES8’s vehicle had a black screen, crashes, door handles could not pop out, or even the network was disconnected. Jianghu people called it “semi-finished products”- The hardware is good, but the software is extremely imperfect. The early ES8 vehicle system also needs to be upgraded online and offline before it can have the OTA remote update capability.

The one that pushed ES8 software issues to the top was the “Chang’an Street Upgrade” incident. When a lady tested the Weilai ES8, she performed a software upgrade operation, and then the entire car had a black screen that could not be operated, and it was eventually blocked on Chang’an Street.

“Semi-finished products” have been one of the main topics that people have been bombarding Weilai for a long time. Fortunately, through OTAs over and over, Weilai gradually fills in the pits of the software.

Thinking caused by ID.3 software defect

Although ID.3’s software problem was before delivery, it was enough to make us realize:

1. In the hardware (Building a car) , the public is very innocent, and the troubles in software will not be worse than Weilai There are fewer such new forces in Chinese car construction;

2. The development of on-board software is actually not as easy as expected, but there are certain thresholds ;

3. As Dith puts it, the public does have a scale effect, but when the troublesome software problem spreads to tens of thousands of ID.3, the trouble index will be doubled. Especially when it can’t be solved by OTA.

By the way, we also experienced Volkswagen ID.3’s car-machine system at the Guangzhou Auto Show. The design philosophy and UI are both good, but the freeze is a bit serious.

ID.3 The news that software defects are exposed can actually reflect a lot of things, such as traditional car giants doing software research and development. What are the advantages? What are the disadvantages?

The advantage is that Volkswagen can share resources within the group through a set of standard software platforms and dilute the cost of software on each vehicle through scale effects. In the long run, this will be beneficial to the future development of the public.

But software development is different from a traditional car with such a long cycle. It requires timeliness and efficiency, finds bugs, can be filled as soon as possible, and iterates new features quickly in response to demand. Take Xiaomi in the mobile phone industry as an example, they basically do a small software iterative upgrade once a week. If you think this example is not appropriate, let’s take Tesla, the smartest one now, as an example. They basically iterate the entire software system once every few months, and some versions have even shorter iteration intervals. Just in the past two days, Tesla has upgraded two versions, and the new smart call function after the update has been used for fun.

This is the advantage of a new car manufacturer like Tesla. It is small in size and strong in execution ability, and can timely respond to user needs.Make adjustments. And This is exactly the weak link of the traditional car companies: it is difficult for the ship to turn around.

During so many years of evolution, the soundness of multi-level decision-making procedures has ensured that Volkswagen is making as few mistakes as possible during the journey of the ship. On the other hand, the automotive industry has entered a “software era” with a faster pace. The improvement of software capabilities is particularly important, and it is necessary to respond to adjustments as soon as possible. In other words, this set of rules of Volkswagen is no longer applicable, and how to make a balance is a problem that traditional car companies like Volkswagen need to think about.

We might as well look at this issue from the perspective of “people”: the engineers of traditional car companies have long been accustomed to their respective duties. Everyone swept the door, and they can be counted as being in their areas of responsibility. Experts, the division of labor is far better than Internet companies. But the negative effect is that if you jump out of your field, you will change from an expert to a little white.

This may be a good model in the field of automotive research and development, but it is extremely dangerous for software research and development. It does not need and should not be cut too finely. At the same time, R & D participants need to have stronger collaboration capabilities and more comprehensive skills, knowledge reserves, and even user thinking. In addition, you must have the ability to quickly find and solve problems.

So, the competition for software talents must be another “invisible battlefield” for traditional car companies. Unfortunately, there are actually very few people who know both software and cars.

From another perspective, at present, only companies with a mass like Volkswagen have the capital to define the matching of software and hardware. It may be difficult for a small car manufacturer to do this because the resources and energy required are too great. In the long run, the automotive architecture will begin to upgrade, and individual ECUs will gradually be replaced by domain controllers, and computing power will gradually be integrated. The experience is better, so that various functions can be matched and matched (such as many linkages between entertainment systems and power systems) , and has the ability to continuously iterate.

Car companies without software self-development capabilities will be uncomfortable in the future. Car company executives, seeing here, is your brain faint?

This article is from the WeChat public account: Geek Car (ID: GeekCar) , author: Yang Jianwen, Liu laugh when