AB Dynamics employees drive a demonstration of the Advanced Vehicle Driving Simulator (AVDS).
Milton Keynes, England:
Autonomous vehicle (AV) startups have actually raised tens of billions of dollars based on promises to develop self-driving cars, but industry executives and experts say robots need remote human observers to help drivers get into trouble. may be required permanently.
The central premise of autonomous vehicles – that computers and artificial intelligence will dramatically reduce accidents caused by human error – has inspired much research and investment.
But there’s a catch: Creating robotic cars that can drive more safely than people is much harder because self-driving software systems simply lack the ability for humans to predict and assess risk, especially When faced with unexpected events or “edge cases”.
“Well, my question would be, ‘Why?'” said Kyle Vogt, CEO of Cruise, a unit of General Motors, when asked if he could see a point where remote human observers would be removed from operations. needed.
Vogt said, “I can bring peace to my clients knowing that there is always a person on hand to help when they need it.” “I don’t know why I ever want to get rid of it.”
This is the first time Cruise has acknowledged the long-term need for remote human operators.
Alphabet Inc’s Waymo and Argo, backed by Ford Motor Company and Volkswagen AG, declined to comment when asked the same question.
GM this month recalled and updated software in 80 Cruise self-driving vehicles after two people were injured in a June crash in San Francisco. US safety regulators said the recalled software could “mispredict” an oncoming vehicle’s path, and Cruise said the unusual scenario would not recur after the update.
For some, the idea that human observers may be here to stay raises more skepticism about the technology.
Truly autonomous vehicles are far behind the optimistic rollout schedule anticipated a few years ago.
In 2018, GM sought US government approval for a fully autonomous car without a steering wheel, brakes or accelerator pedals that would enter its commercial ride-sharing fleet in 2019. That vehicle, the Cruise Origin, is no longer ready to begin production until spring. 2023, Vogt said.
In 2019, Tesla Inc. CEO Elon Musk promised one million robotaxis “for sure” next year – although his company’s “full self-driving” offering has been criticized because its cars drive themselves without a human behind the wheel. Are not able to steer and are ready to take manual control in case of emergency.
In a June interview on YouTube, Musk said that developing self-driving cars was “way harder than I originally thought.” But when asked about timing, he said Tesla could make it “this year.”
Tesla did not respond to a request for comment for this story.
The unrelenting promise of true autonomy has raised the stakes for the AV industry.
“If these companies don’t succeed in the next two years, they won’t exist,” said Mike Wagner, CEO of Edge Case Research. “It’s a matter of keeping or keeping quiet at this point.”
remote humans watching
Many AV startups today use humans as remote observers, along with safety drivers sitting behind the wheel.
Those remote humans are an added expense, but help self-driving cars handle edge cases. These can include something as basic as an unfamiliar set of lane closures during road construction, or erratic, unpredictable behavior by pedestrians or human drivers.
When a robot driver encounters an edge case, “it puts its hands up and says, ‘I don’t know what’s going on,'” said Kosha Kawe, CEO of Imperium Drive, which runs remote operators for cars. As is using humans. The English city of Milton Keynes. Over time, those people would act as “air traffic controllers,” overseeing the growing number of autonomous cars.
Cruise’s Vogt says the company’s AVs on the streets in San Francisco currently rely on humans for less than 1% of the time. But in the hundreds, thousands or even millions of AVs, that would add a significant amount of time stopping on the road waiting for human guidance.
Imperium Drive’s Kawe said that as more self-driving cars — which are more predictable than humans — hit the roads, the number of edge cases will go down, “but you’ll never get zero edge cases.”
“Decades from now you still won’t find 100% truly autonomous vehicles,” Kawe said.
Nevertheless, competition is increasing. Some Chinese cities are pushing to allow active AV testing more quickly.
The need to tackle edge cases and cut costs for everything from sensors to the number of humans to hit the market has also intensified as investor funding for autonomous cars dwindles.
Doubts have arisen in the minds of investors as to how quickly the autonomous business will become profitable. Simple or slow AV, low-speed routes such as trucks or last-mile delivery services on highways or sets are likely to reach profitability first, but it will still take years to get there.
Overall investment in future mobility startups has slowed, according to investor website PitchBook, with AV-focused companies being hit particularly hard, representing less than 10% of venture investments in the second quarter. (Graphic: https://tmsnrt.rs/3Rzy04y)
Mobility tech funding decline in 2022 https://graphics.reuters.com/AUTOS-AUTONOMOUS/TECHNOLOGY/egvbkradkpq/chart.png
Investments in AV startups declined to $958 million in the quarter. According to Pitchbook, just two years ago AV investing was booming, as Alphabet’s Waymo raised $3 billion, Didi’s AV unit for $500 million and Amazon.com Inc. acquired AV startup Zoox for $1.3 billion. .
‘Rush to Market’
Autonomous systems aren’t as capable as people because their “perception and prediction algorithms aren’t as good as how a human brain processes and makes decisions,” said Chris Boroni-Bird, an independent consultant who previously worked at GM and Waymo. advanced-vehicle programs.
For example, a human when seeing a ball rolling down the road — harmless in itself — would assume that one of its children could chase and hit the brakes much quicker than Evie, Boroni-Bird said. .
“I worry that AV companies will hit the market without proving safety as better than human-powered vehicles,” he said.
The problem is that there are “tens of billions of potential edge cases” that AVs could face, said James Roth, CEO of AB Dynamics, which runs tests and simulations on cars, including advanced driver-assistance systems (ADAS). Foundations of autonomous driving features.
Auto data startup Viejo Group Ltd. receives 18 billion data points daily from millions of connected cars and is helping with simulations for AVs, said Sarah Larner, executive vice president of strategy and innovation.
“But there are so many variables like weather, you can take an edge case and then have to layer in all the different forms,” she said. “That’s actually millions of outputs.”
driverless delivery
In its track tests for cars, AB Dynamics uses a robotic arm that plans to retrofit slow-moving mining and agricultural trucks to make them largely autonomous.
Roth envisions a remote team of humans monitoring a fleet of self-driving mining trucks operating in closed environments.
He doesn’t see that the scenario is working for vehicles in faster, more open environments because it can be difficult for remote human observers to react adequately to threats.
Within the next 12 months, British online food delivery and technology company Okado Group plc will partner with autonomous vehicle software startup Oxbotica to build a small fleet of driverless delivery vehicles – backed by remote human observers – that ply just a few streets on designated routes. Will Work. Never drive more than 30 miles (48 km) per hour in a small town in the UK.
“At 30 mph, if a vehicle panics, it can hit the emergency brake and ask for help,” said Alex Harvey, Okado’s chief of advanced technology. “It seems like a very viable strategy at low speeds.”
“But you can’t play that game on the motorway,” Harvey said, because hard stops in edge cases would pose a safety risk.
Harvey said it should take about five years for Okado to develop a profitable driverless delivery system. He said more than half of Okado’s UK customers can’t drive over 40 mph with the AV. Eventually, the service could be rolled out to Ocado customers such as US retail chain Kroger Co.
(Except for the title, this story has not been edited by NDTV staff and is published from a syndicated feed.)