Reduce costs and improve travel efficiency. Let's talk about the development trend of Baidu's driverless cars.

In order to let everyone experience cutting-edge AI technology and learn new AI knowledge, Baidu held the first "AI, Let's Go" AI Open Day event on July 16. The event focused on shared driverless cars, with the theme of "Driving is too tiring, taxis are too expensive? It's time for Robotaxi to take action." Xu Baoqiang, director of Baidu's autonomous driving product operations department, shared Baidu's autonomous driving technology and the development trend of shared driverless cars.

Xu Baoqiang, Director of Baidu's Autonomous Driving Product Operations Department, shared his thoughts on the event

The rapid development and implementation of artificial intelligence technology has brought new possibilities to many traditional industries, including the automobile industry, which has been developing for more than a hundred years. At this stage, the development of the autonomous driving industry presents a situation of competition between China and the United States. The United States started earlier in this field, while Chinese companies have shown a strong trend of catching up. China's urban road scenes are more complex, and the autonomous driving technology trained in this scene has stronger capabilities. In addition, China's advantages in infrastructure such as vehicle-road collaboration and 5G cloud driving are also very obvious, which can further accelerate the maturity of domestic autonomous driving technology.

Guidehouse continues to rank Baidu as one of the top four "leaders" in the global autonomous driving field

The industry generally believes that from lagging behind in the era of fuel vehicles to competing against each other in the era of electric vehicles, in the era of intelligent vehicles, domestic autonomous driving technology and vehicles are fully capable of surpassing foreign countries. In 2020 and 2021, Guidehouse, a world-renowned research organization, listed Baidu as one of the four "leaders" in the global autonomous driving field, and it is the only Chinese company in the leader camp.

5G cloud driving

5G cloud driving connects remote operation to the real car through the 5G network, and the low latency of the 5G network can be used to remotely operate the vehicle. The vehicle's 360-degree camera will display the captured images in real time on the large screen at the rear. At the same time, the various states of the vehicle will also be displayed on the large screen, making it convenient for the "cloud driving" to take over the vehicle in time. Of course, the "cloud driving" plays a more important role as a safety officer, because Baidu Apollo's autonomous driving can easily cope with complex urban roads.

Baidu officially launched the research and development of autonomous driving technology in 2013 and released the Apollo open platform in 2017. To date, Baidu Apollo's autonomous driving technology has covered multiple fields such as public transportation, sanitation, retail, sightseeing, logistics, agriculture, education, mining trucks, etc., covering scenarios such as transporting people, delivering goods, and special work. Among them, the most technically difficult, most relevant to daily life, and with the greatest commercial potential is L4-level urban road shared travel.

Baidu Apollo Environmental Perception

Baidu's autonomous driving technology is already fully capable of handling L4-level shared travel scenarios on urban roads, just like having an "AI veteran driver" driving for you. To train this "AI veteran driver", you first need to rely on data-driven core algorithms, which are equivalent to its brain; secondly, you need to hone your technology through road tests; and finally, you need to provide "eyes in the sky" assistance through vehicle-road collaboration.

Regarding the issue of autonomous driving safety that users are most concerned about, Baidu Apollo's autonomous driving has designed multiple safety guarantees and fault redundancy from software to hardware, and will also ensure its effectiveness through a detailed testing system from simulation to closed test sites and then to actual road tests. At the same time, every driverless car on the road will pass the driving license test of the regulatory agency. So far, Baidu Apollo's autonomous driving has obtained 293 test road licenses nationwide, including the highest level and most complex Beijing T4 level license, which is still exclusive to date. Xu Baoqiang said that Baidu's autonomous driving car currently has a success rate of 99.99% on urban roads, and its comprehensive safety is higher than that of human drivers.

To achieve L4-level shared travel on urban roads, in addition to powerful autonomous driving technology, excellent vehicles are also needed. In June this year, Baidu launched the fifth-generation shared driverless car Apollo Moon. Compared with previous generations of products, the sensor layout of Apollo Moon is no different from that of mass-produced cars on the market. At the same time, its hardware reliability is further improved, and the core computing unit and sensors have a 5-year warranty, which can achieve 20,000 hours of trouble-free stable operation.

In addition to these product design advances, Apollo Moon has also achieved an industry-leading, unprecedented low cost of only 480,000 yuan, which is about one-third of the industry's general autonomous driving vehicles. The low cost of the vehicle means faster large-scale operations, allowing autonomous driving services to enter our lives faster.

Since 2019, Baidu has opened shared driverless car travel services in Changsha, Cangzhou, Beijing and other cities. Users can experience shared driverless car travel services in these areas through Baidu Maps or Apollo Go App. As of the end of 2020, Apollo Go has served more than 210,000 passengers. In the next three years, Baidu will expand shared driverless car services to 30 cities, deploy more than 3,000 shared driverless cars, and serve more than 3 million users.

Driverless shuttle bus