A fatal accident involving a pedestrian and one of Uber's self-driving vehicles could have been prevented, a new report claims. An employee warned th
A fatal accident involving a pedestrian and one of Uber‘s self-driving vehicles could have been prevented, a new report claims.
An employee warned the ride-sharing giant that there were issues with Uber’s autonomous-driving technology just days before Elaine Herzberg, a 49-year-old Arizona woman, was struck and killed.
The email, which was sent to several high-level executives at Uber, warned that the self-driving cars had been involved in several accidents, likely due to ‘poor behavior of the operator of the AV technology,’ according to the Information.
Scroll down for video
A fatal accident involving a pedestrian and one of Uber’s self-driving vehicles could have been prevented, as a new report claims the firm was warned of safety issues ahead of the crash
Robbie Miller, a manager in the testing-operations group, sent the email on March 13th.
The crash, which involved Herzberg being hit by a manned autonomous vehicle in Tempe, Arizona, occurred just five days later on March 18th.
Miller sent the email to seven of Uber’s attorneys and executives, including Eric Meyhofer, who leads the firm’s autonomous vehicle division.
Among the concerns laid out in the letter were that routine accidents involving Uber’s autonomous vehicles frequently went unchecked, while human backup drivers often didn’t receive proper training and hadn’t all been thoroughly vetted.
In the email, Miller says ‘poor behavior of the operator’ was usually the cause of the accidents, while ‘several of the drivers appear to not have been properly vetted or trained.’
An Uber SUV after hitting a woman on March 18, 2018, in Tempe, Ariz. A new report claims an Uber executive warned others of safety issues just five days before the crash took place
He then urges Uber to begin using two backup drivers in each vehicle.
It was later revealed that the car’s backup safety driver in the fatal accident was watching The Voice using Hulu on her phone when the crash occurred.
Miller also notes that Uber’s self-driving vehicles had been ‘hitting things nearly every 15,000 miles,’ while some vehicles had been damaged ‘nearly every other day’ in February 2018.
Meanwhile, vehicles narrowly avoided accidents as much as every 100 miles, as backup drivers were forced to control the vehicle as frequently as every one to three miles, the Verge noted.
And when collisions did occur, backup drivers were rarely fired, Miller alleged in the letter.
In the email, Miller references one accident in particular, where an Uber swerved off the road and onto the sidewalk and continued to drive there.
HOW DID AN UBER AUTONOMOUS VEHICLE HIT AND KILL A WOMAN IN ARIZONA?
A self-driving Uber vehicle struck and killed a pedestrian in the first death involving a fully autonomous test vehicle on March 19, 2018.
The accident prompted the ride-hailing company to suspend road-testing of such cars in the US and Canada.
The Volvo SUV was in self-driving mode with a human back-up operator behind the wheel in Tempe when a woman walking a bicycle was hit.
Elaine Herzberg, 49, died in hospital.
Police have said that the victim, 49 year old Elaine Herzberg, stepped out in front of the car suddenly and they do not believe the car was to blame.
Uber suspended its self-driving vehicle testing in the Phoenix area, Pittsburgh, San Francisco and Toronto.
The testing has been going on for months as car makers and technology companies compete to be the first with cars that operate on their own.
Uber’s self-driving car crash that led to the death of a mother-of-two could have been avoided, driverless vehicle experts have claimed.
Cortica, a firm that develops artificial intelligence for autonomous vehicles, has analysed the dash cam video.
The company concludes the car, which failed to brake or swerve before the collision, had enough time to react and potentially save Ms Herzberg’s life.
Speaking to CNET, Cortica’s CEO Igal Raichelgauz said the firm’s self-driving AI system detected Ms Herzberg 0.9 seconds before impact.
At this point the car was around 50 feet (15 metres) away.
He said the autonomous car’s cameras and radar system should have had enough time to pick up the pedestrian and react to the situation.
Driverless cars are fitted with a system of cameras, radar and lidar sensors that allow them to ‘see’ their surroundings and detect traffic, pedestrians and other objects.
An AI computer system then decides what actions the car takes to avoid a collision – a setup that is supposed to work as well at night as during the day.
A top executive for the maker of Lidar sensors used on Uber’s self-driving car said she was ‘baffled’ as to why the vehicle failed to recognise Ms Herzberg.
The case was ‘essentially ignored’ for days, Miller said.
He also argued that Uber should downsize its self-driving car fleet to prevent further safety issues – an idea that was ignored from Uber’s high-level executives.
What’s worse, Miller also argues that Uber was too slow to respond to reported incidents within its self-driving vehicle fleet.
His email warning of the safety concerns never received a response from any of the executives it was sent to and Miller eventually left the company three days after he sent the email.
Uber maintains that it continues to focus on the safety of its self-driving cars.
National Transportation Safety Board (NTSB) investigators examine the self-driving Uber vehicle involved in the fatal accident. The firm has since resumed its self-driving car testing
‘The entire team is focused on safely and responsibly returning to the road in self-driving mode,’ the company told the Information.
‘We have every confidence in the work that the team is doing to get us there.’
Uber temporarily shut down its self-driving car operations after the crash occurred in March.
Earlier this month, however, the firm announced it would resume testing on roads near its offices in Pittsburgh.
It comes as a New York Times report found that Uber hasn’t resolved many of the safety issues with its self-driving cars.