On the first anniversary of Baidu’s car-making, Jidu used SIMUCar and the core of wisdom to build the first car robot

On January 11, 2021, Baidu, as the world’s leading artificial intelligence platform company, officially announced that it will enter the automotive industry as a vehicle manufacturer and form a smart car company. Geely Holding Group became a strategic partner of the new company. Subsequently, Baidu used AI algorithms to name the new company more than 100 names. After several rounds of screening, the name was finally determined as “Jidu”, and its brand interpretation was “the culmination of Baidu’s AI capabilities”. More than a month later, Jidu was officially established, with Xia Yiping, the former head of Ford Motor’s Intelligent Vehicle Association and Mobike’s co-founder and CTO, as CEO.

Jidu, a start-up technology company that inherits Baidu’s powerful artificial intelligence and Internet technology genes and has Baidu Apollo’s leading autonomous driving capabilities, will use Baidu Group’s 8-year experience in the field of automotive intelligence to reshape intelligence. The form of automobile products, leading or even changing a new era of intelligent travel? It has become the first urgent problem to be solved in front of CEO Xia Yiping.
Also in the same year, Xiaomi Lei Jun announced after a wave of intensive research that his last venture would be “ALL IN” to build a car. Huawei, with thick eyebrows and big eyes, also “betrayed” his original promise to lead the manufacturing after the “soul theory”. Car, 2021 has become the first year for technology companies to build cars. Who will be the top technology company leaders who have “rolled” to the extreme this year? At present, Baidu’s car-making brand – Jidu is the “leader”.
Breaking through the pattern – building the first mass-produced carRobot
“We are not building cars or electric vehicles in the traditional sense,” Xia Yiping, CEO of Jidu, told the visiting media with an imaginative metaphor on June 1, 2021, “The future of unmanned vehicles or the future The smart car is actually a “four-wheeled Robot” in essence. Then the forward-looking concept of “automotive Robot” was presented at the 2021 Baidu CREATE World Conference and announced by Baidu head Robin Li.
Jidu’s automotive robots run through the three major product concepts of “free movement, natural communication, and self-growth”.First, the car robot has the L4 level of autonomous driving ability and can move freely; second, the precise recognition of human-vehicle interaction and speech semantics allows the car and the human to “naturally communicate”, understand the user’s emotions, and respond in real time.applicationThird, car robots can learn and iterate themselves according to the user’s usage habits, and continuously optimize the experience of autonomous driving and smart cockpits to achieve “self-growth”.
It is different from the functions of traditional cars and the definitions of new forces. It perfectly integrates Baidu’s end-to-end core AI technology and capabilities with Apollo robotaxi’s autonomous driving capabilities, and the most advanced Geely Haohan SEA platform as a mechanical blessing. A new concept, a new category, and a new product were born. It is committed to the research and development and popularization of the world’s top autonomous driving and human-computer interaction technologies, creating revolutionary automotive robots with emotion and intelligence, and giving people more space and time to realize more possibilities.
With the improvement of the level of intelligence, the complexity of software has changed, especially the basic capabilities of the underlying software have undergone qualitative changes. The core of vehicle technology and engineering is shifting from the traditional hardware level to the software level. At the same time, the traditional model of hardware research and development first, and then software research and development can no longer meet the needs of intelligent vehicles in the new era, because the real vehicle development and verification time left to the software team is very short, and it is good to ensure that there are fewer bugs before SOP. , and no energy to develop new features.
Faced with such industry-level problems, Xia Yiping, the appointed head of Baidu Group, subversively pioneered the automotive research and development of SIMUCar (Software Integration Mule Car) with the forward-looking “software first” thinking logic. The model has subverted the car manufacturing process of traditional car companies. To put it simply, SIMUCar simulates a vehicle electronic and electrical architecture and SOA environment that can be used for software research and development, and provides a supporting system that integrates L4-level driving automation capabilities. It can be understood as a neural network for future automotive robots. Based on this neural network, software and hardware decoupling development can be realized, so that the engineers can carry out software mass production development in the simulated vehicle environment in advance, and continue to optimize and iterate. While the software is iterating, the hardware is also optimized based on the vehicle development process. Simultaneous software and hardware iteration and optimization greatly improve the development efficiency of smart cars.
At present, based on SIMUCar’s research and development model, Jidu’s “automotive robot development process” has efficiently completed the intelligent driving research and development goal, and completed the industry-leading high-speed and urban dual-domain intelligent driving integration only 10 months after its establishment, which has been verified in advance. The mass production of L4 autonomous driving capability is safe and reliable. This means that the “brain” of the Jidu automotive robot has initially possessed the ability to “freely move”, and is fully preparing for mass production and listing in 2023. It can fully provide users with a leading, safe and reliable autonomous driving experience when it is delivered. It is not provided by OTA after delivery.
Following the signing of the contract with Continental Group and the deep chat with Bosch Group, in the face of the intelligence of the future automotive core battlefield, Jidu CEO Xia Yiping chose Qualcomm and NVIDIA to undertake the intelligent driving and intelligent cabin system jointly developed by Jidu and Baidu. When asked why he used the industry’s most advanced chips instead of current products, Xia Yiping’s answer once again showed his double thinking as an engineer and product manager.
At the business level, Jidu builds automotive robots. No matter in terms of product design, product concept, software interaction, including the ability to achieve intelligence, it is based on the “forward development” of technological products and meets the needs of future users. Think logically. Therefore, chips such as 8295 and Orin are selected to carry the high computing power required by high-level intelligent cabins and intelligent driving.
For consumers, Jidu’s car robots will implement many functions according to the user’s needs for future travel modes, such as high-level autonomous driving, human-computer interaction and entertainment in the cockpit, etc. Therefore, Jidu needs the most advanced vehicles. The high computing power required for these functions can be carried by standard-level chips, so as to perfectly realize the functions recognized and expected by these users.
Therefore, whether it is Continental, Bosch, Qualcomm, or NVIDIA, I believe that Chidu’s automotive robots will use the capabilities of leading suppliers in these industries to maximize “intelligence” and make future cars intelligent, interconnected, safe and reliable. Presented in front of concentrated users.
Fast and efficient – 2023, delivering the first automotive robot that defines the highest level of intelligence in Chinese cars
From the 18-member team that was officially registered and established on March 2, to the 800-member large-scale organization that began to take shape in 2022; from the first exposure of the first clay model photo on June 10 to the first visit of Li Yanhong on July 7 a month later. Participate in confirming the design and shape of the first product; from the signing of the German Continental Group on August 6 to the signing of Qualcomm and NVIDIA respectively on November 29 and early 2022; from August 30, the first full-size clay model of the first product was air-driven for the first time. From the completion of the experimental test to September 24th, Jidu has carried out the development of SIMUCar software to simulate the intelligent driving of prototype vehicles, and then to December 31st, the latest development of Jidu’s automotive robots, SIMUCar, has successfully integrated high-speed and urban dual-domain intelligent driving. The one-year anniversary of Baidu’s car-making, or 2021, the collection is efficient and fast. So what kind of car robot will be unveiled at this year’s Beijing Auto Show in 2022? Can it become the industry benchmark for smart cars after mass production is delivered in 2023? Can Xia Yiping use his forward-looking thinking and actions to become a disruptor in the smart car industry? I believe the market will soon be able to give an answer.

The Links:   3HAC030923-001 3HAC037682-001

Published on 08/28/2022