Saturday, June 19, 2021
Home > Opinion > Uber’s Self-Driving Pilot Program

Uber’s Self-Driving Pilot Program

National Transportation Safety Board officials analyze the Uber self-driving car that killed Mrs. Herzberg.
Voiced by Amazon Polly

Uber has ambitiously pushed their self-driving pilot program since its first deployment of AVs (Autonomous Vehicles) on public roads in Pittsburgh, Pennsylvania, in 2016. Their main incentive has been to provide safer, cheaper, and more convenient commute on public roads. According to the National Highway Traffic Safety Administration, cars kill 103 people on US roads daily and 94% of car accidents are caused by human error. Uber believes that these figures can be drastically decreased hrough the application of self-driving vehicles. However, they still face many obstacles before they can deploy AVs on a large-scale.

First announcements of Uber’s intention to deploy AVs were in February 2015, around the same time Google announced its interest in AVs. Uber was welcomed by Pittsburgh Mayor Bill Peduto who claimed that this innovation was a necessary step for Pittsburgh to “be a 21st-century laboratory for technology” and was given relative freedom to access Pittsburgh’s roads. Researchers from Carnegie Mellon University were hired, and soon, the first AVs were rolled out. Efforts were then made by Uber to test automated cars in other locations, such as San Francisco, California, Tempe, Arizona, as well as Toronto, Canada. Within half a year after Uber kicked off their self-driving pilot program in Tempe, Arizona, a fatal car accident occurred resulting in the death of 49-year-old Elaine Herzberg. This was the first death of a pedestrian caused by collision from an AV and Uber has already suspended all testing of AVs as they work with authorities and the US National Transportation Safety Board to investigate the incident.

Who is at fault

Defenders of Uber and self-driving technologies claim that there is nothing to be concerned about, citing that thirty to forty thousand people die each year from human-driven automobile accidents. One death from AVs should only be expected. However, Bryant Walker Smith, a law professor at the USC, believes that, “this incident was uncomfortably soon in the history of automated driving. In the United States, there’s about one fatality for every 100 million vehicle miles traveled, and automated vehicles are nowhere close to reaching this many real-world miles. This arguably first fatality may not tell us much statistically, but neither is it reassuring.” He along with others believe that this accident could have been avoided. The human safety driver – someone sitting at the wheel of the vehicle – should have been able to manually take over and stop the vehicle. Also, the latest self-driving models have only one LiDAR (the system that scans the vehicle’s surroundings) compared to the seven attached to previous vehicles. According to engineers who built the system, “the singular LiDAR array has a known blind spot.” The New York Times reported that Uber’s AVs can barely travel above thirteen miles without human assistance, while in comparison, Waymo’s AVs average five thousand six hundred miles. Uber’s vehicles may have not yet been ready for public roads. They seem to have been pushing the limits of their program beyond what they were able to manage. This is most likely the effect of their staggering budget as well as competition with other companies. These factors drove Uber to take risks that they believed could allow them to deploy AVs on a massive-scale first – an accomplishment worth a potential $7 trillion as proven by a study conducted by Intel. There was the Uber vs Waymo lawsuit, where Waymo, the Google self-driving project, claimed that a former Google employee downloaded 14,000 files of confidential data on AVs before leaving to found Otto, another self-driving vehicle company. The Google employee later became one of the leaders of Uber’s self-driving pilot program after Uber purchased Otto. Upon noticing that Uber’s LiDAR circuit board looked precisely the same as theirs, Waymo sued Uber for patent infringement and stealing trade secrets in February 2017. The trial was settled, with Uber giving Waymo an equity stake of $245 million. Uber currently generates a negative profit with a loss of money in the billions per year.


Uber’s net revenue (blue) vs net losses (red)

Founder and former CEO of Uber Travis Kalanick has also claimed its self-driving pilot program to be basically unachievable and that it would most likely reduce fares by 80%.

The Future of Uber

Uber’s current CEO Dara Khoshrowshahi publicly announced on the 9th of May this year that they plan to resume self-driving tests within a few months. However, this could prove to be quite difficult. Of the four locations that Uber has previously negotiated with, Toronto, Tempe, San Francisco, and Pittsburgh, only Toronto still has a favorable relationship with the ridesharing company.

Ever since Uber’s DMV (Department of Motor Vehicles) permit expired on March 31, the company has lost access to testing on state roads. Even after two months have passed, Uber has not attempted to renew their permit saying that they are currently “heads down” in the incident at Arizona. They’ve also announced that their AVs “would not operate in the state in the immediate future.”

Arizona governor Doug Ducey, someone who once eagerly welcomed Uber’s self-driving cars, has now denied all tests of AVs in public roads. “As governor, my top priority is public safety. Improving public safety has always been the emphasis of Arizona’s approach to autonomous vehicle testing, and my expectation is that public safety is also the top priority for all who operate this technology in the state of Arizona… [this accident] is an unquestionable failure to comply with this expectation,” he said.

Uber has already pulled out of Arizona and announced that they will not attempt to test self-driving cars there again. Meanwhile, two other companies – Waymo and Cruise – continue their tests in the state.

Bill Peduto was under the impression that Uber would not only bring innovation but also an increase in jobs as well as free AV rides, however, these expectations were not met. This being the case, he has laid out strict conditions for Uber for when they want to return to Pittsburgh for testing. He required that before Uber returns, a full federal investigation be completed (usually takes 12-18 months) and that Uber’s AVs must not exceed 25 mph.

Despite problems in Pittsburgh, Uber decided to expand to Toronto and open a new self-driving center there. The relationship between Toronto and Uber is relatively well maintained. Uber has dedicated themselves to supporting the Vector Institute, an AI research organization founded by Raquel Urtasun – the head of Toronto-based research for self-driving cars.

It is unlikely that Uber will try to start testing in a new location due to their reputation plummeting after they broke the ‘undeserved’ trust of both Doug Ducey as well as Bill Peduto. Both initially welcomed and supported Uber’s self-driving efforts, however, Uber has failed to meet their expectations.

Uber is falling behind competing companies. As Uber waits for an investigative report and registration to testing ground, GM, Waymo, and Tesla are all making significant progress with their AVs. As Quartz News reports, “General Motors’ Cruise Automation will launch a car without a steering wheel next year. Waymo’s autonomous vehicles in the Phoenix metropolitan area (which includes Tempe) are picking up passengers. Tesla, despite enduring a fatal accident involving Tesla’s Autopilot in the US, has upgraded its safety functions rather than scale back its technology (driver error was found to be partially at fault).”

For their self-driving pilot program to exceed competitors, Uber will have to rebuild relationships with state officials and upgrade their vehicles with more safety measures so that no major incident is repeated. Otherwise, it would most likely be necessary for them to shut down this program and focus to another.

Future of AVs

Even at the rapid pace that companies are pursuing AVs at, it may still take a while before AVs become the new norm. It is much harder for people to forgive casualties caused by an AI compared to one caused by another person. Also, the majority of people are still concerned about being the passenger of a self-driving vehicle.


Survey conducted by researchers of University of Michigan Transportation Research Institute on motorists

The leading reasons for why they would not ride in a driverless car are as follows:


These concerns from the public will have to be addressed before any company can roll-out massive quantities of driverless cars.





Loading Likes...

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.