Early today self-driving cars made headlines when an Uber vehicle in “autonomous” mode struck and killed a pedestrian outside of the crosswalk. Following the incident, Uber pulled all of its self-driving cars from the road:
Uber has removed its self-driving cars from the roads following what is believed to be the first fatality involving a fully autonomous car.
“The vehicle involved is one of Uber’s self-driving vehicles,” the Tempe police said in a statement. “It was in autonomous mode at the time of the collision, with a vehicle operator behind the wheel.”
Autonomous mode means the car is driving on its own. During tests, a person sits behind the wheel as a safeguard.
While outrage following incidents like these is to be expected, self-driving cars will be back on the road for further testing before the greater public even realizes it. And like it or not, these vehicles will begin to replace “manually” operated automobiles sooner rather than later. And when their replacement of traditional vehicles becomes commonplace, many drivers will likely find yourself “driving” one, for one simple reason:
The autonomous car doesn’t need to be perfect, it merely needs to be a “better” driver than the (average) human. Once that technology exists, cost as well as economies of scale will make it very prohibitive and/or expensive to drive traditionally-operated cars.
It is worth starting by sharing some basic auto accident statistics in the US, courtesy of ASIRT:
Annual United States Road Crash Statistics
- Over 37,000 people die in road crashes each year
- An additional 2.35 million are injured or disabled
- Over 1,600 children under 15 years of age die each year
- Nearly 8,000 people are killed in crashes involving drivers ages 16-20
- Road crashes cost the U.S. $230.6 billion per year, or an average of $820 per person
- Road crashes are the single greatest annual cause of death of healthy U.S. citizens traveling abroad
These are astoundingly high casualty rates, far higher than firearm fatalities. Some analysis of road accidents from Statista would likely show a great deal of these accidents are caused by drivers who are: drunk and/or substance impaired, tired and/or sleep-deprived, reckless, or distracted, just to name a few.
While it will take lots of time and testing, with accidents like today’s occurring along the way, if a self-driving car can be developed that is merely as good on average as a sober, attentive, capable driver… that is good enough to reduce auto fatalities substantially.
If you are drunk and/or tired, and need to get somewhere, would you rather drive, or let the car drive you? The choice should be obvious.
And as the performance of self-driving cars improves, their benefits will increase, and their costs will decrease. Cars have had cruise control for years, and are now being built with automatic emergency braking, lane keeping assist, and some Tesla vehicles can operate without any driver intervention for short periods of time. This fact is not lost on insurance companies, who insure less risky cars at lower rates.
When a fully autonomous car is available that is statistically “safer” than a manual car, they will be insured at a far lower rate. One must also consider how lawmakers will react, as they will likely dole out incentives for “safer” self-driving vehicle purchases, as well as penalties for manual operation. If the self-driving car is better on average than the human, the effects will be exacerbated, and auto accidents will likely go down substantially as a result.
This will not be lost on auto makers, who will convert their production lines to mass-produce self-driving cars. Cars that can only be manually driven will be produced less frequently, and at a higher cost, as self-driving cars become mass marketed and sold to consumers. One day, vehicles without a self-driving option may even become so expensive to buy, operate and insure that they become a novelty item.
Who will want to purchase a traditionally-operated car when the cost of doing so is significantly higher than purchasing a self-driving one? While many will be wary of ceding a significant amount of control to a machine (or tech company, or government), and others will simply prefer to drive themselves, price and economics have a funny way of determining behavior, and that is before human safety is added to the equation.
In addition to cost and safety, the convenience of a self-driving car cannot be understated. Inebriation and exhaustion pales in comparison to youth in this department. Drivers under the age of 25 are statistically far more likely to be involved in auto accidents, which is reflected in the “price premium” this age group pays to insure and/or rent a car. Driver licenses in the US are generally not issued until age 16 at the earliest. But children still need to be driven to soccer practice, which requires an adult licensed driver.
That will be less and less necessary as automation becomes more advanced. Eventually, a self-driving car will be automated to the point where kids too young to legally drive can put themselves in the car, choose the destination, and “drive” themselves. Some have even theorized that this will result in more cars on the road, or that auto ownership itself will become far lower, with the service being “rented” by major providers (such as Uber). Whether or not that is the case is up for debate, but the practicality of a self-driving car on families with children below the legal age to drive cannot be understated, notwithstanding the risk and expense of insuring a younger, riskier driver.
In spite of today’s crash, self-driving cars will resume testing in short order. More accidents will happen, and more headlines and outrage will follow them. And while many (including myself) may not be completely comfortable with the thought of self-driving cars, they are coming, like it or not, for one simple reason…
…we will truly be “only human” in comparison to self-driving cars once the technology reaches its full potential.
Update: While the investigation is still in its preliminary stages, it appears that the self-driving Uber vehicle was not “at fault” in the accident:
In the first hint at the investigation being carried out by Tempe police, the San Francisco Chronicle reported that the Tempe police chief said her preliminary investigation suggested that Uber wasn’t at fault. Police Chief Sylvia Moir described the victim, the possibly homeless 49-year-old Elaine Herzberg, as pushing a bicycle laden with plastic shopping bags when she abruptly stepped from the center media into a lane of traffic before being struck by the car.
“I suspect preliminarily it appears that the Uber would likely not be at fault in this accident, either,” Moir said.
Moir added that she “wouldn’t rule out” the possibility of charges against the backup driver in the vehicle, even though she said it appeared that neither a human driver or an autonomous car could’ve reasonably been expected to avoid the victim, who was caught on video abruptly stepping into the roadway into oncoming traffic.
It appears at least in this instance, the self-driving car was no better or worse than the human driver. Expect road testing to begin again soon.