Baidu launches nighttime driverless taxis to China

Baidu, a Chinese internet giant best known for its search engines and autonomous driving, is making big strides.

The robotaxis will be available to the public in Wuhan from 7 am to 11 pm, with no safety drivers. It was previously only possible to operate the unmanned vehicles in Wuhan between 9 AM and 5 PM. This updated plan will cover one million customers in Wuhan, which is home to more than 10,000,000 people.

Baidu, like most autonomous vehicle startup, uses a combination of radars, cameras, and lidars to aid its cars in low-visibility situations. This contrasts with Tesla’s vision-based solution.

In AugustBaidu launched fully driverless robotaxi rides in Q3, charging passengers taxi-rates. Apollo Go, the robotaxi-hailing app of the firm, launched in Q3. completed more than 474,000 ridesThis is an increase of 311% year-over-year. Apollo Go had received 1.4 million orders in total as of Q3.

This sounds like a huge revenue stream for Baidu. But one must be careful and ask the question: How many of these trips can you subsidize with discounts? How many of these are daily, repeatable routes than novelty rides that are only available to early adopters? It’s not unusual for Chinese robotaxi operators to recruit the public to ride their vehicles in order to boost performance.

It can be difficult to discern which of China’s robotaxi start-ups are leading at this stage. Their expansion is dependent on their relationship with the local city They operate in certain areas, and large cities have the power to pass local legislations.

Autonomous driving is one of few consumer internet segments that has room for growth. It is receiving warm support from all levels of government. Wuhan, a central China industrial city, was the first to allow robotaxis to drive the public, without the need for in-car safety officers. Now, even at low-light nights, it seems that the city is comfortable with driverless cars.

Baidu has put in a lot to help self-driving cars arrive sooner, but that is just a small amount of skeptical. Baidu’s visual-language model, which can identify rare or unseen objects in long-tail situations, is one of its moats. Wenxin is the same large model that supports its AI. text-to-image art platform.

Baidu said that “The model will allow autonomous vehicles the ability to quickly make sense an unseen item, such as special vehicle (firetruck or ambulance) recognition and plastic bag misdetection. explained. “In addition, Baidu’s autonomous driving perception model—a sub-model of the WenXin Big Model—leveraging more than 1 billion parameters, is able to dramatically improve the generalization potential of autonomous driving perception.”