Nerd For Tech
Published in

Nerd For Tech

Depth丨 Autonomous Driving Industry in 2021 — Part 2

The Truth About Mass Production: From Fancy Concept to Pragmatic Safety

The automotive industry has never been short of unbridled creativity, and fully self-driving cars may be the “peak” of the imagination. However, since the industry has developed to the present, L5-level autonomous driving is still far away. When the whole industry has to land, what autonomous driving is growing into in “mass production” will be clear to the public.

Presently, the self-driving products and services that have been or are being mass-produced mainly have three similarities.

1. The mass production of L4 autonomous driving, mainly commercial vehicles, is concentrated in the fields of buses, logistics, distribution vehicles, passenger car freight, and other fields.

In the field of public travel, WeRide, which focuses on L4 autonomous driving technology, has begun to explore multi-scenario services such as public transportation and intra-city freight. Zoox, Baidu Apollo, and QCraft have also successively deployed Robobus tracks. In the field of unmanned delivery, Haomo.AI, an autonomous driving AI company independent from the intelligent driving prospective branch of the Great Wall Technology Center, and the L4-level unmanned delivery vehicle “Xiaomanlu” produced in cooperation with Alibaba Dharma Academy have achieved large-scale mass production delivery. In the field of freight, after TuSimple rang the first bell of the global launch of self-driving trucks, Inceptio also announced the realization of L3 mass production, and the subsequent OTA can be upgraded to L4. Pony.ai and PlusAI have also announced the time for the mass production of self-driving trucks.

In general, L4-level autonomous driving is a high-level, leap-forward technology, with great technical challenges and difficulties. While commercial vehicles with relatively simple road conditions, low accident rates, and significant improvement in quality and efficiency are more likely to achieve large-scale implementation of L4.

According to CITIC Securities, the potential market space for self-driving software and hardware for commercial vehicles is about 59.18 billion dollars, and the service space for autonomous driving operations is about 0.44 trillion dollars. Players who have won a crucial position in this field have promising business prospects.

2. The mass production of autonomous passenger vehicles will be based on L2-level technologies, and the exploration of L4 and higher-level passenger would be slowed down.

In the field of passenger cars, there has been a misunderstanding about “autonomous driving” or “assisted driving”, which has triggered people’s careful thinking about leapfrog technology. We find that most of the self-driving passenger cars that can be mass-produced are L2-based. For example, Xpeng Motors has always defined self-driving as an L2-level auxiliary system; In an email to the California Department of Motor Vehicles (DMV), Tesla also admitted that FSD, like Autopilot, is a Level 2 autonomous driving system. The penetration rate of L2-level autonomous driving of Great Wall Motors is as high as 40% as well.

In summary, in order to develop reliable technology and achieve mass production, and meet the growing demand for new energy vehicles, the vehicle manufacturers such as Tesla, Xpeng, Great Wall, and technology companies like Haomo.AI incubated in OEMs all choose incremental development, gradually evolving from L2 and L3 assisted driving to L4 and L5 unmanned driving.

At the 3rd BAAI Conference, the top conference in the AI field in 2021, Gu Weihao — CEO of Haomo.AI, mentioned that autonomous driving will be improved with the development of computing power, data, network, and the laws and regulations.It will be implemented gradually and orderly according to the three layers of “from low speed to high, carrying things to carrying people, and business to civilian use”.

Behind Autonomous Driving: High-quality Labeled Data

The three essential elements for artificial intelligence to operate are computing power, algorithms, and data. Together, they form the whole of artificial intelligence.

Among these three elements, computing power is the ability of technical facilities, the algorithm is the working method, and data is the basis for optimizing the algorithm. In other words, the first two are equipment and capabilities. Data is the knowledge material, which plays an important role. High-quality training data will maximize the efficiency of artificial intelligence, while low-quality AI data will be not only impossible to improve efficiency, but also will hinder the evolution of artificial intelligence to a certain extent.

End

Outsource your data labeling tasks to ByteBridge, you can get the high-quality ML training datasets cheaper and faster!

  • Free Trial Without Credit Card: you can get your sample result in a fast turnaround, check the output, and give feedback directly to our project manager.
  • 100% Human Validated
  • Transparent & Standard Pricing: clear pricing is available(labor cost included)

Why not have a try?

source:https://www.163.com/dy/article/GVCKRLFT053823M4.html

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
ByteBridge

ByteBridge

Data labeling outsourced service: get your ML training datasets cheaper and faster!— https://bytebridge.io/#/