Waymo and Tesla use remote human assistance to support self-driving systems
Despite advanced AI, self-driving cars still need humans to step in when technology struggles
Alphabet’s Waymo and Tesla's recent findings have revealed that self-driving cars are not fully autonomous yet, and humans remain crucial to keeping them safe. Both companies have shared information about their “remote assistance” services, which enable trained operators to support vehicles when software faces any technical confusion or dangerous scenarios.
Waymo, which provides robot taxi services in ten US cities and is set to expand globally, has confirmed that it has 70 remote assistants who monitor 3,000 vehicles. Half of these assistants are contractors in the Philippines.
According to the company, most of the driving situations are resolved by its autonomous system independently, with human assistance utilised only in a small number of edge cases.
Waymo highlights that all their assistants must complete US traffic rule training while holding driving licences and passing drug and alcohol assessments. Waymo Vice President and Global Head of Operations Ryan McNamara stated that the company employs a US-based team to manage its most difficult challenges, involving accidents and interaction with police.
On the other hand, Tesla uses a different method to approach its business operations. The company operates two domestic offices of remote operators, located in Austin and the San Francisco Bay Area.
According to Tesla AI technical program manager Duy Cao, all Tesla operators need to work from US locations while passing background checks and maintaining active driver licences, which he stated in his California Public Utilities Commission filing.
Tesla has decreased the number of human monitors who observe drivers inside vehicles; however, certain cars still require either chase vehicles or operators to maintain driver oversight.
Moreover, Carnegie Mellon University autonomous vehicle researcher Philip Koopman points out that even highly advanced AI systems cannot predict all possible scenarios on the road. “For the foreseeable future, there will be people who have a role to play in the behaviour of the vehicles, and therefore a role to play in safety,” he said.
-
What happens if ChatGPT gains access to your financial accounts? Experts are alarmed
-
Anthropic seeks legal pause on Pentagon supply-chain risk decision: Here’s why
-
'AI washing' or real shift? Atlassian cuts 1,600 jobs in latest tech shake-up
-
Experts predict AI will trigger biggest shift in mathematics history
-
China’s cyber agency raises concerns over OpenClaw AI
-
WhatsApp plans major change for younger users
-
Musk unveils Tesla, xAI joint project ‘Macrohard’ amid advanced AI push
-
Nvidia secures $2 billion deal with AI cloud provider Nebius
