Google, apparently hoping the turn the screws on federal regulators so that it can continue to speed development of its autonomous driving technology, was to appear before the Senate Commerce Committee today, where Chris Urmson, head of the Googlemobile program, was to tell senators to speed things up. Google is a subsidiary of the Alphabet Companies.
Urmson was set to ask Congress to give regulators authority to get autonomous wheels on the ground quickly, even as regulators dealt the autonomous driver program a delay late last week. The National Highway Traffic Safety Administration (NHTSA), even though it had already defined software-based control systems as “drivers” last month, said Friday that there huge hurdles to overcome before the autonomous technology can move ahead.
Urmson’s appearance before the committee today seems like an attempt to circumvent regulatory action so that the Googlemobile program can move ahead quickly. Indeed, Reuters, quoting his prepared text, noted that Urmson was there to help get fully autonomous vehicles on the road.
“We propose that Congress move swiftly to provide the secretary of transportation with new authority to approve life-saving safety innovations. This new authority would permit the development of innovative safety technologies that meet or exceed the level of safety required by existing federal safety standards, while ensuring a prompt and transparent process,” Urmson’s text said. What his remarks fail to clarify is that currently vehicles must meet 75 safety standards before they are approved. Apparently, Google believes that it’s software- and hardware-based solution which has now failed in a crash with a city transit bus is far better than standards the search engine wants to set aside in its quest to be first.
Google is just the tip of the iceberg. Other automakers like GM and Ford, which has established major programs or have purchased subsidiaries to develop autonomous vehicle technology, and technology companies like Uber, are in the race of a generation to be the ones that produce and sell self-driving cars, a market that doesn’t – and may never – exist. Current studies show that most people are frightened of self-driving technology. In fact, the same studies show that drivers prefer to keep their hands on the wheel.
Google’s view could very well be a case of computer-driven hubris on the part of some firms who believe that they have the proper vision for the future and, until we buy into their vision, everyone is plainly wrong. What irks developers most now is that conflicting state and federal rules are hamstringing their efforts to put autonomous technology on the roads next to the driving public. The companies complain that rules such as the ones proposed by California in December which require human controls and a driver in every car – not bad safety checks – impede the testing and development of self-driving vehicles.
Google, headquartered in California, is hugely disappointed by the state’s action. “If every state is left to go its own way without a unified approach, operating self-driving cars across state boundaries would be an unworkable solution and one that will significantly hinder .. the eventual deployment of autonomous vehicles,” Urmson emphasized before the committee. Urmson, in fact, believes that many federal safety standards are, in the face of superior self-driving technology, useless. For example, he maintains in his testimony that autonomous technology would end the need for a rearview mirror and other requirements though there seems to be no mention of just how autonomous tech would do this.
Some wonder whether Urmson’s rather quick appearance before the Senate panel may be Google’s way of forcing the issue before regulators show that autonomous tech may not be all that it is proposed to be. Driving this suspicion is an accident last month in Mountain View, Calif., where a Lexus 350h, part of Google’s vaunted, crash-free test fleet, smashed into a city bus. It seems that Google’s software engine decided that it was the bus’ turn to yield for the Googlemobile as it sought to enter traffic. Perhaps it was because the software didn’t recognize a simple item like yielding the right of way that the vehicle made the decision (not the driver) to challenge the much larger municipal vehicle, with the almost-inevitable result, an accident.
Google’s test mule, traveling quite slowly and expecting the bigger bus to yield struck the side of the bus. There was no report of injury, and no agency has yet assessed fault. For a bit, it looked as though Google might have succeeded with its ploy of accepting minimal responsibility for the crash, while heaping blame on the oncoming vehicle. However, that plan failed when NHTSA Administrator Mark Rosekind Friday said the agency would look into the accident. Google’s stance, by the way, in the wake of the crash is that it has fixed the software glitch that allowed the accident in the first place, and now its fleet of self-driving vehicles is set to go and face oncoming traffic.
In all of Google’s maneuvering, it is not that it hasn’t achieved a degree of success. In January, for example, NHTSA said it might waive some rules to allow more driverless cars on the roads as part of an effort to move things ahead. Also, NHTSA said software-controlled intelligent engines could be considered drivers for testing purposes, lowering another barrier.
Back-tracking, the agency scaled the race to the autonomous car back considerably Friday when it recognized that while it would be easy to approve a self-driving vehicle that met current rules, there are significant legal hurdles to overcome before cars could appear without steering wheels. NHTSA, at the same time, said it would be writing new standards for self-driving cars within six months, not the result Google had been hoping for. Indeed, it was to tell the Senate today that new rules are needed soon as it wants to offer fully autonomous vehicles on the roads quickly.
Information for this story came from Reuters and Automotive News.