Focus on Autopilot Driving -Tesla Spends $4 Billion to Purchase Samsung Cameras!

Whether it is a camera or other sensors such as lidar, the purpose of the industry is very unified, all of which are for the automatic assisted driving system to be closer and closer to the real automatic driving, and also to lay the foundation for the future of the car. According to the The Korea Economic Daily, Samsung Electric Machinery Co., an electronic component subsidiary of SamSung group, has won a contract worth about 4 trillion to 5 trillion won (about US $3.2 billion to US $ 4billion) to supply new camera modules for most of Tesla's electric vehicles.

In the next few years, Samsung's new cameras will be used in all of Tesla's vehicles, including the Model 3 sedan, Model S sedan, Model X SUV, Model Y crossover, Cybertruck and Tesla Semi semi truck, according to people familiar with the matter. It is reported that Samsung was also awarded a contract last year to supply cameras exclusively for the Tesla Cybertruck. It is reported that Samsung will start mass production of the new camera as soon as next month.

Major upgrade to Tesla's assisted driving system camera hardware capabilities

As mentioned earlier, Musk has publicly stated many times before that the automatic assisted driving of the camera solution is the optimal solution in Musk’s personal opinion. Since Tesla launched the autopilot system in 2014, it has been working on camera-based vision fusion automatic. The development of the assisted driving system. At present, Tesla's fully automatic driving FSD system has been actively developed and promoted. The main purpose of the cooperation between Tesla and Samsung components is that Tesla wants to seek a clearer camera to advance the clarity in the automatic assisted driving system, and help the operation and judgment of the automatic assisted driving system to be more clear and agile.

According to reports, Samsung is providing a 5-megapixel camera module of Samsung version 4.0. According to official propaganda, the resolution of the version 4.0 camera is 5 times higher than that of the previous generation version 3.0. The conclusion of this cooperation will also be a major upgrade of the camera hardware capabilities used in Tesla’s current automatic driver assistance system sensor suite. According to previous Tesla car hardware data, the previous Tesla driver assistance system camera resolution The rate is only 1.2 million pixels, which is indeed a huge improvement.

Tesla and Samsung will definitely build a lasting partnership in the future

According to reports, the reason why Samsung was able to establish a cooperative relationship with Tesla this time has also opened a competitive mode within the industry. According to reports, Samsung finally defeated LG to win all the orders for Tesla’s cameras. According to people familiar with the matter, in the next few years, Samsung's new camera will be used in all of Tesla's car products, including Tesla model 3, model Y, model X, model S, as well as future Tesla pickups Cybertruck and Tesla Semi truck model; according to reports, Samsung not only won a $4 billion contract from Tesla in 2022, but in 2021, it has already won a camera contract specifically for the Tesla pickup Cybertruck. It can be said that Tesla and Samsung will certainly establish a very lasting cooperative relationship in the future. 

Is it possible without lidar?

Different users may have their own understanding of this problem. Even industry practitioners have different ideas about the development and route of technology. At present, some enterprises have always used lidar as the main sensor of future automatic assisted driving systems. Although it will also be equipped with cameras inside, but more of it is to assist lidar; Tesla in the automatic assisted driving system, more use cameras to collect surrounding data in real time, with 2D pictures as data support, through many cameras The images collected in real time are used to calculate the surrounding environment and vehicle models, so that the operation of the automatic assisted driving system can be carried out in real time.

At the current time node, some car enthusiasts have experienced Xiaopeng NGP and Tesla FSD. The sensors and algorithms of the two models are also completely different. However, according to the experience of car enthusiasts on high-speed sections, the automatic driving assistance systems of the two enterprises are very top-level in terms of user experience optimization, whether it is lane centering assistance, automatic lane change or even overtaking, The stability and availability are very high. Therefore, at present, it is still in the so-called l2+ assisted driving system in the industry. Although there are differences in sensors, in terms of experience and perception of users, it can be considered as the same.

In the future, the automatic assisted driving system will be gradually innovated and upgraded. The author believes that the fusion of more sensors will definitely be better than the single acquisition and modeling of the camera, which is more conducive to the stability and upper limit of the system..

Using the time flight calculation method, combined with real-time peripheral rapid data collection, can effectively help the auto assisted driving system to more comprehensively perceive the surrounding environment, which is more conducive to highlighting the ability of assisted driving; Although the lidar has a strong ability to perceive the surrounding environment, because the collected data and later calculation are relatively difficult, and the price of the lidar itself is relatively high, the high cost of the lidar and the cost of the high computing power chip must be much higher than that of the camera head. In addition, it does not mean that the lidar must have better functional performance. What is important is that the enterprise should optimize the full stack self-developed algorithm, Only the combination of software and hardware can give the best experience to the automatic driving assistance system.

Write at the end

According to recent media reports, Tesla has applied to the Federal Communications Commission (FCC) for the use of a new radar, and I can't guess what medicine Musk is selling in the gourd. After the launch of Tesla's autopilot in 2014, it brought a huge tremor to the industry, and it also made more auto companies witness the importance of innovation, and then began to cause many companies to follow in Tesla's footsteps and start to exert force.

Looking at the world on the shoulders of giants can quickly follow the development of the industry, but if in the rapid development, it can also create new thinking strategies and directions, so that you can also become a new giant, which will certainly bring great changes to yourself and the industry. The author also expects that in the era of electric transformation and intelligent upgrading, global automobile enterprises can continue to explore and innovate.