MADISON – Uber Technologies grounded all of its driverless cars deployed in pilot programs in Tempe, Arizona, Pittsburgh and San Francisco, after a crash Friday in Arizona.
Reportedly, the Uber SUV was on a main road in the city when another vehicle failed to yield while turning left in front of it. The vehicles collided, the Uber vehicle rolled on to its side.
(Source: Fresco News)
A spokeswoman at Uber told EE Times, "We are continuing to look into this incident and can confirm we had no backseat passengers in the vehicle."
Here’s what Uber has so far confirmed.
- Uber’s vehicle was in self-driving mode at the time of the crash.
- There are no known serious injuries at this time.
- Two vehicle operators were in the car’s front seats. (Uber says this is standard for the company’s self-driving pilot.)
- The Uber car was carrying no passengers at the time of the crash. (Currently, Uber has a passenger pilot both in Pittsburgh and Tempe.)
- In San Francisco, Uber recently received a permit from the California DMV to operate two of its robo-cars. Currently, those cars are “in development mode only” and not picking up passengers, according to Uber.
Josie Montenegro, a spokeswoman for the Tempe Police Department, told Reuters that the accident occurred when the driver of a second vehicle "failed to yield" to the Uber vehicle while making a turn. She added, "The vehicles collided, causing the autonomous vehicle to roll onto its side.”
With no further information available, it appears that the Uber vehicle was not responsible for the incident.
Phil Magney, founder and principal advisor for Vision Systems Intelligence (VSI), at first blush, sees it as “an unfortunate coincidence that the car at fault happened to collide with an Uber test vehicle.” He added, “Obviously Uber came upon another ‘edge case’ and reinforces the need to collect tons of data before full automation can be realized.”
Questions to ask
It’s logical to dismiss this as an accident caused by a human-driven car. It’s just as logical, especially in light of the safety promises made by the promoters of self-driving cars, that Uber’s robo-car, loaded with crash-preventive sensors, proved itself defenseless in the face of an ill-timed maneuver by a careless human driver.
But first, there are a few questions that must be asked and answered.
Mike Demler, senior analyst at The Linley Group, asked a question that goes to the heart of the matter: Where was the Uber SUV (Volvo) exactly hit?
The accident photo doesn’t seem to tell the whole story. Rather, it is confusing. Let’s assume that the local NBC report is correct. It reported that “according to Tempe police, it was headed south on McClintock Drive when another vehicle turned left in front of the Uber at Don Carlos Drive.They crashed and the Uber rolled onto its side.”
But the photo is puzzling. Demler asked: “How could the Volvo end up on its right side pointing in the opposite direction?”
So, Demler’s questions to the police are:
- If the Volvo was heading South when it was hit, had it just turned in that direction from Don Carlos, or had it proceeded through the intersection heading South on McClintock?
- The Volvo clearly wasn’t hit from the front. Was it hit from the rear and spun around, or was it a side-to-side collision?
Demler also asks Uber: “Will you release your data from the on-board self-driving system?”
He said, “I want to know where the Volvo was hit to know if its 360-degree images could have picked up the Ford SUV in time to respond. That’s the only technology ‘inside’ the car that could have possibly helped.”
Demler added. “If it was hit from the rear, there’s not much the Volvo could have done. Even if it saw the Ford coming from the rear, it still couldn’t accelerate out of the way. Maybe all that damage to the Ford’s hindquarter happened after the Volvo spun around and bounced off it and the car behind it." Further, he speculated: "But – if the Volvo was traveling in the middle lane and the Ford came from its side, it might have been able to make an evasive maneuver into the right lane."
Next page: Handover