Latest News

Nvidia and car suppliers form partnerships to revive the self-driving drive

AI, a network of partnerships, and tech suppliers including Nvidia, are the key to accelerating progress in the self-driving industry.

However, many automakers are still unsure. They are concerned about the high costs and scalability but also want to know whether there is enough demand from customers to justify an expensive bet.

The advent of vehicles that can drive themselves would revolutionize the world of transportation. However, making this technology safe enough for public roads is proving to be more difficult and expensive than anticipated.

While some companies, such as Alphabet Waymo or Tesla, have chosen to develop their own autonomous vehicles in-house, others such as General Motors or Ford Motors have given up on this. AWS and German supplier Aumovio made an announcement at the CES in Las Vegas, this week. Kodiak AI, a company that makes autonomous trucks, and Bosch announced a partnership to increase production of autonomous trucking sensors and hardware. Nvidia, a company that makes AI chips, has released its "next-generation" platform. This platform will be used by Lucid Group and Uber in their robotaxi alliance. Mercedes-Benz announced this week that it would launch an advanced driver-assistance program in the United States this year, powered by Nvidia chips. The system will allow its vehicles to operate autonomously within city streets while under driver supervision.

Artificial intelligence, the driving force behind self-driving cars, is now proving to be a powerful tool for development. It could help reduce costs.

AI and generative AI are a "big accelerator" for the industry, "because they allow... significant development and validation while using significantly less resources," Ozgur Tohumcu said. He is general manager of Amazon's cloud division Amazon Web Services. Western automakers also face pressure to keep up with China's drive to be the leader in autonomous driving development and adoption. Last month, China approved two cars that have Level 3 autonomy, which allows for hands-free driving. The auto industry has defined five levels for autonomous driving. From 'cruise control' at Level 1, to fully self-driving without the need of a human driver at Level 5, there are five levels.

Jochen Hanebeck is the CEO of German chipmaker Infineon. He cautioned against "market fantasies" that fully autonomous cars would be commonplace in a few short years.

He said that rather than invest in new investments for fully autonomous vehicles, automakers prefer to use driver assistance technologies, also known as Level 2 technology, which is already available, but requires constant attention from drivers.

Hanebeck stated, "I do not see a tsunami moving towards Level 5." There have been a number of announcements of robotaxi deployments in China, Europe, and the Middle East. But Jeremy McClain of Aumovio’s autonomous mobility unit said that to expand the areas covered by these robotaxis, more data is needed, as well as fleets and logistics.

"MAKES US FEELS LIKE WE ARE THERE"

Self-driving cars are a hot topic. Elon Musk, Tesla's CEO, promised that the electric car maker would have one million self-driving vehicles on the roads by the end of 2019. Tesla only launched its small robotaxi service last year, six long years after Musk made his bold prediction.

It was a problem that self-driving cars are susceptible to a multitude of unexpected events, also known as "edge cases." These can be easily misunderstood by the vehicles. Experts often cite the example of a human driver slowing down when they see a ball roll into the street because a child might be chasing it. But a self driving car would only react if it saw the child. Ford and GM, among others, abandoned their money-losing auto-pilot units after the first bubble burst. The death of GM Cruise was accelerated after it dragged a pedestrian for 20 feet (6 meters) and struck him.

Ali Kani is the general manager of Nvidia's automotive team. He said AI enabled advancements to address key flaws in self-driving technologies.

Kani says that there are some pieces of foundational technology which make him feel as if he's in the field.

Morgan Stanley analysts in a CES note said that Tesla was still years ahead, despite Nvidia’s new Alpamayo self-driving platform giving legacy carmakers a boost and helping them to pressure Tesla. Many in the industry believe that Nvidia's open-source platform, which is a good place for Tesla competitors to gather, will give them a leg up.

Russell Ong said, "In a way, it was almost like Apple and Android," referring to Tesla’s proprietary system versus Nvidia’s decision to release Alpamayo, an open-source version. Reporting by Abhirup in Las Vegas, Akash in Bengaluru and Nick Carey. Editing and proofreading by Peter Henderson, Matthew Lewis and Matthew Lewis.

(source: Reuters)