Colorado State Senator Owen Hill (R-Colorado Springs) has introduced Senate Bill 213, a four-page bill that would actively permit self-driving cars on Colorado’s roadways.
The driverless car industry is rapidly growing, and a recent report
predicts that there will be 10 million self-driving cars on American roadways by the year 2020. States around the country are competing for this growing industry by trying to make autopilot research and development in their state more appealing.
We’re clarifying that, like any other technology such as ABS brakes or adaptive cruise control, if you can follow the rules of the road you don’t need any additional permission to operate here ... Colorado could be the Silicon Valley of all kinds of vehicle technology on this front.” - State Senator Owen Hill
Compared to the bill, Colorado’s Department of Transportation is using nearly identical language to govern driverless cars in the state. However, the industry has been pushing for more active acceptance – concrete permission to operate driverless cars – as opposed to passive acceptance that can be revoked at any time.
Supporters estimate that adoption of driverless cars would decrease traffic fatalities and reduce congestion. More than 600 people died on Colorado’s roads in 2016 and the legislation estimates that driverless cars could reduce crashes and fatalities by up to 90 percent.
The bill's advocates also argue that testing the systems in Colorado would bring hundreds of new jobs to the state.
However, opponents of the technology also point specifically to the issue of jobs. Bringing a few hundred Research & Development jobs to Colorado will do little to balance out the tens, if not hundreds,
of thousands of Coloradans who could become unemployed as a result of this technology.
Census data shows that one of the most common professions in Colorado is "Truck Driver."
Across the United States, at least five million people drive professionally. These include hospitality drivers, truck drivers, and delivery workers, among others. This doesn't include the people who drive part-time for ride-sharing services like Lyft and Uber. Both companies tout having 1 million drivers, but the number of active unique ride-share drivers is likely in the hundreds of thousands. Millions of other Americans work in industries supporting commercial drivers.
Currently, there still needs to be a human driver behind the wheel of a self-driving car. One firm in Europe is developing software to create automated caravans, where one human drives the lead truck and other automated vehicles follow his movements. Ultimately, the goal of the technology is to enable truly autonomous travel, which would eliminate entire types of jobs.
While self-driving cars on the road may be inevitable, opponents argue that it is an entirely different matter to hasten the technology’s advancement. They say that it is illogical for the government to encourage autonomous vehicle research without creating a safety net for the citizens who rely on human-driven vehicles to make a living.
This includes, as an example, repair and auto-body shops.
Over six years of testing, Google has reported just 11 minor accidents during its autonomous vehicle testing. None of these were the fault of the self-driving cars, according to Google's self-reported statistics.
If the 90 percent crash reduction statistic is correct, that would mean a massive reduction of work in the auto-repair industry. While even self-driving cars need maintenance, crash reduction alone could result in almost 500,000 Americans losing their jobs, according to the Bureau of Labor Statistics.
Again, if vehicular automation is truly the way of the future, then the State of Colorado must take steps now to prepare to accommodate the tens of thousands of Coloradans whose jobs are threatened by the technology.
In addition to potential job losses, safety and reliability of self-driving cars is another area of concern.
Anytime software is introduced to an otherwise manual process, the likelihood for non-user generated error always increases. This is always outweighed by the benefits that the automation brings. The possibility of a computer crash, as an example, is outweighed by the ability to do complex calculations in fractions of a second.
The consequence of computer failure is amplified when introduced into an autonomous motor vehicle.
In June of 2016, a driver was killed when his Tesla Model S autopilot feature mistook an oncoming tractor-trailer for an overhead road sign. Because the car's radar software misidentified the truck as “signage,” the autopilot never applied the breaks or corrected itself when the car began drifting across the centerline. And since the truck was hidden “against a brightly lit sky,” the driver never saw it coming either.
Even if the driver was unable to see the oncoming truck, there is no way of knowing whether he would have also drifted into the truck's path if he was driving.
Add in the prospect of sabotage and the risk of automated cars becomes even greater.
Recently, the organization Wikileaks released a trove of Central Intelligence Agency documents detailing digital surveillance techniques known as Vault 7
. One CIA program from 2014
focused on developing the ability to remotely control a vehicle’s computer systems. While this program focused on hacking into computers within manually controlled vehicles, it is easy to see how this strategy could be applied to autonomous vehicles.
Some have speculated that this could create a nearly undetectable way to assassinate someone. Conspiracies aside, the risk of autonomous vehicles being hacked or sabotaged is very real.
Hacking an autonomous vehicle puts human lives at risk in ways that typical computer hacking does not. In 2015, researchers Charlie Miller and Chris Valasek demonstrated
the ability to hack into the Wi-Fi system in a Jeep Cherokee. While they were able to adjust the volume and turn on the windshield wipers, they were also able to get into the vehicle’s transmission system and bring the SUV to a halt in the middle of the highway.
Luckily, this was part of a pre-planned experiment and no one was harmed. It shows the potential for sabotage as more and more vehicle functions are taken out of the driver’s hands and entrusted to software.
Like any form of automation, Coloradans must decide whether the benefits of self-driving cars outweigh the risks.
Creating new types of jobs in the automated car industry means endangering jobs that many Coloradans rely on. The average truck driver or delivery driver isn’t going to be able to transition into a career as a software engineer developing AI for self-driving cars.
Reducing the number of automobile crashes
could also mean shuttering auto-body shops across the state.
It is entirely possible that Coloradans will see these negatives as 'collateral damage' necessary for technological advancement. But the debate needs to happen.
While State Senator Owen Hill’s legislation does an excellent job catering to the interests of those who support self-driving cars, it does nothing to allay the real concerns of tens of thousands of Coloradans whose livelihoods are threatened by the technology's rapid advancement.