This Is Why Tesla’s Robotaxi Launch Needed Human Babysitters

“It’s a demo or test using the protection driver – it’s not a [autonomous vehicle] Deployed, “Bryant Walker Smith, a law professor at the University of South Carolina, said that he studied autonomous vehicles.

Blooper and sensor

Tesla’s technology has kept quiet about many details. And it is hard to reach specific conclusions about its technology from social media posts uploaded by the riders. However, a few of these posts show less rides. In a video, a robotuxi is trying to turn a left A double yellow line is seen to be crossed In the coming traffic. On the other, a robotuxi Obviously a UPS truck stopped and failed to identify the opposite in the parkAnd the front seat protection monitor has to interfere with the car stopping.

A Robotaxi “Phantom Breaking” uploaded a YouTube video-is coming to a stop with no apparent reason — an incident that tesla has also flagged hundreds of users Federal Government investigatedThe In contrast to actual self-driving technology, users need to keep an eye on the streets of complete self-driving.

According to Tesla’s website, the service breaks for bad weather. One YouTube Had their journey off for a rainfall; Robotuxi dropped the rider in an Austin Park when the wind began to strap them around them. Minutes, according to A videoThe same robotuxi picked up the Creator to continue their journey. However a poster that opposed the above Reported The cars perform “innocent” in heavy rain.

The initial blupers are not surprising, experts say. A human driver is needed to interfere with complete self-driving (supervising), and it seems to be the same now that Robotuxi is the same as Philip Copman at the University of Carnegie Melon, a study by autonomous vehicle protection. He says that the slip-ups made by robotaxis are no different from what human drivers do on the road. However, the quality of autonomy is considered protection, so it refers to the videos – and the “rough edges” of the technology – make people nervous.

Camera Quander

The launch has reopened public debates about a core tenet of Tesla technology: it is to use the camera alone to realize it as well as drive and “decision -making”. The musk and its company have long argued that artificial intelligence complemented by the camera collected by the camera is enough to manage a safe, driverless car. CEO has promised that all its cars can become autonomous without any changes, with a simple push of updated software (though Tesla Renewal On this claim). Other companies are viewing more expensive sensors, including radar and leaders, as important validators and support. (Leader has dramatically decreased in prices; many Chinese automaker are now incorporating sensors Every car they sell.)

Progress in big language models In the auto industry, some people have confirmed that the musk method is correct. In a podcast interview published this week, Kyle Vogot, a former CEO of General Motors AV Unit Cruise, Argument Multiple vehicles-mounted cameras plus advanced models images “truly correct” can be. (Enjoy Cruise One of its driverless vehicles is hit by a pedestrian after one. The company was not transparent with the controllers about this event, Later a report was found.)

For Cummings, reports outside Austin have confirmed to his belief that the camera alone is not enough to autonomously drive. “There is no robotic system that exists that criticize the protection – which means people can die [using it]”It has ever been successful in a single sensor strain,” he said. “Why Tesla thinks they can do what they have never done before.”

A metric that can reveal internal success in Tesla: how fast it is. Kasturi boldly said in May that Tesla had thousands of thousands – and perhaps one million on the road next year – autonomous vehicles. The company seems to be inspired. Ay Job postingTesla is recruiting for additional vehicles operators, who are paid for driving around Austin to collect data. But of course the musk is not a stranger Unmeatable the deadlineThe