Car Infotainment Distractions Can Be Deadly

A study just released by the AAA Foundation for Traffic Safety shows that many 2017-model cars have infotainment features that can dangerously distract drivers.  Such distractions may be an important reason why, after declining for years, the annual car-fatality rate in the U. S. rose in 2016.

A report in the Washington Post describes how University of Utah researcher David Strayer led a study involving about 30 different 2017-model vehicles, ranging from a Ford F250s to a Tesla Model S.  Test subjects performed infotainment-feature-related  tasks while being monitored in various ways, and while also having to push a button every time a buzzer went off.  The delay between the buzzer and the button-pushing has been found to be a good indication of how distracted the driver is.  And while all this was going on, the subjects were driving down a low-traffic residential street, so the whole experiment was conducted during real-life driving.

The findings were not encouraging.  Some tasks, such as programming a navigation system, took an average of 40 seconds to do. Some vehicles were twice as demanding as others on average, based on a rating called the "overall demand," which included a variety of visual and cognitive tasks. 

The Post article quotes AAA chief executive Marshall Doney as saying that there are some things we have no business doing while behind the wheel.  The U. S. National Highway Traffic Safety Administration issued guidelines to automakers in 2012, asking them to block certain kinds of tasks involving infotainment systems unless the vehicle is parked.  But this call has gone largely unheeded.

It's interesting that the two vehicles with the most demanding electronics were the Honda Ridgeline, Honda's answer to US-brand luxury pickups, and the Volvo Inscription, another luxury passenger car.  Among the cars with the less-demanding systems were the Ford F-250 pickup and the Chevrolet Equinox.  It begins to look like this problem is a side effect of carmakers' attempts to load up their vehicles with many-featured electronics in the higher-end models especially.  The practice of piling on new software features, whether useful or not, is familiar to anyone who has used computers recently, which basically means everybody. 

It's one thing to sit at your desk and fume at Excel for burying a function you liked in an avalanche of newer features that you find largely useless.  But it's quite another thing to be driving on an LA freeway and trying to program your next real-estate appointment address into your GPS system at the same time.  As unwise as such an action is, people will do it, and get away with it too, at least for a while. 

The car manufacturers face a dilemma.  The IT features of a car are sometimes one of the main things that set a vehicle apart from the competition, so you can't expect the carmakers to stop trying to offer newer and more exciting infotainment features. 

But safety can get lost in the shuffle.  A car company can either install lock-out technology that just flat prevents the user from doing time-and-attention-intensive tasks while driving, or the firm can simply warn the consumer not to do such things while in motion.  Most companies have taken the latter course, with the excuse that if people do stupid things, well, we told them not to do that, so you can't blame us. 

In other areas of human endeavor, this approach has been tried and found wanting.  In the field of occupational health and safety, for example, workplaces in factories used to be extremely dangerous, with bare moving belts and moving parts everywhere.  Employees were simply warned to stay out of harm's way, but that wasn't always possible, and a lot of people got killed.  With the advent of workmens' compensation insurance and government supervisory agencies such as the Occupational Health and Safety Administration (OSHA), attitudes changed.  Now employers largely accept the responsibility for building in safety in their plants with shields, safety interlocks, and procedures that approach being foolproof in many cases.  No small part of this change is due to pressure brought to bear on miscreant manufacturers by insurance companies that got tired of paying out premiums to workers injured at needlessly hazardous plants.

It may be that auto insurance companies will need to play a role in making sure that infotainment features don't distract drivers to death.  When you realize that only a certain fraction of cars on the road are replaced with new ones every year, the suspicious upward trend in car fatalities becomes even more ominous.  One wonders what would happen if everybody like me (our newest car is a dozen years old) bought a new car and started trying to use the navigation system in an injudicious way. 

I don't know whether auto insurance rates vary much from vehicle to vehicle based on the model's safety record, but if a prospective buyer learned that the bells-and-whistles luxury model he was about to buy carried a huge insurance price tag, he might hesitate.  And the carmaker might do something about installing a gentle form of lockout, making it at least inconvenient to use some of the more demanding features while actually driving, which would make the insurance companies happier.

What is generally regarded as the first fatality in an autonomous (self-driving) vehicle—the crash that killed Joshua Brown—occurred allegedly while Brown had set his Tesla Model S in self-driving mode and was watching a video.  Admittedly, this is carrying distracted "driving" to the extreme, and was against the manufacturer's instructions.  But it shows that if a system allows the driver to do a stupid thing, somebody somewhere will eventually do it, and sometimes with dire results.

No movie, song, or GPS information is worth a person's life.  Carmakers need to realize that they are undermining a decades-long trend of improved car safety with the fancy gizmos they are shipping with each new vehicle.  If the average consumer isn't smart enough to avoid the new hazards, something else needs to be done.  Voluntary compliance with the 2012 NHTSA lockout guidelines would be nice.  But if history is any guide, automakers may need the encouragement of laws and regulations to implement new electronic infotainment features that are both attractive and safe to use.

Sources:  The Washington Post carried the article by Ashley Halsey III, "New cars have more distracting technology on board than ever before," in the October 3, 2017 online edition at https://www.washingtonpost.com/local/trafficandcommuting/new-cars-have-more-distracting-technology-on-board-than-ever-before/2017/10/04/8dc1e91e-a880-11e7-b3aa-c0e2e1d41e38_story.html.  I also referred to the original AAA report, "Visual and cognitive demands of using in-vehicle infotainment systems," which is available at https://www.aaafoundation.org/sites/default/files/Report_0.pdf.  My blog on the Joshua Brown accident appeared at http://engineeringethicsblog.blogspot.com/2016/07/self-driving-car-fatality-no-1-joshua.html on July 4, 2016.

Can the Digital Future of Cars Save Lives and Time?

Despite all the positive changes the automobile has wrought, there are still a few big problems.  Leading the list is the rate of automotive fatalities and injuries—thousands of people die in car crashes every year, and many times that number are seriously injured.  Next on my list is the millions of person-hours wasted each year by people sitting in slow traffic—needlessly long commute times.  Add the carbon footprint of each car to that picture, and you can see plenty of room for improvement in the way we use machines to get around. 

At a one-day event called AutoMobilityLA held at the annual Los Angeles Auto show that runs through Nov. 27, New York Times reporter Tom Volek surveyed a number of digital technologies that promise to deal with all of these problems.  But as with many nice ideas, the difficulty is how we're going to get from here to there without making things worse before they get better.

Take self-driving cars, for instance.  According to Dr. Alexander Hans, a blogger at a site called www.driverless-future.com, several studies have shown the potential for a self-driving taxi to perform the transportation work of six to ten privately-owned vehicles.  He also claims that the first widespread use of self-driving cars will be in fleets of self-driving taxis operating in restricted geographic areas such as densely populated districts of urban areas (think places like Singapore, where the first commercial self-driving taxi fleet debuted last August). 

Maybe these forecasts are right, but computer simulations leave out certain factors that may be decisive.  For example, there are lots of cabs in Manhattan, and there would be even more if the existing cab companies had not engaged in rent-seeking by restricting the total number of medallions available and fighting innovative unlicensed services such as Uber and Lyft.  But even if all the restrictions on cabs and taxi-like services in Manhattan were removed, I think you would still have a lot of cars clogging the streets, many of them privately owned. 

A city is a complex thing, and it is a mistake to assume everything else will stay the same if all you do is insert a change in the transportation mix.  That is why new freeways get crowded so quickly and the race to alleviate congestion by building more freeways never seems to be won.  Better and more congenial transportation attracts residential and commercial development until the new transportation mode is just as crowded as it used to be, and then people go somewhere else to repeat the cycle.

And even more important than alleviating commuting time and headaches is safety.  We are told that once most cars on the road are self-driving ones, that auto accident rates will plummet.  Given the fact that most auto fatalities are due to operator misjudgments and not mechanical failures, I can believe that.  Computers don't get drunk and try to impress their friends with their alcohol-impaired driving skills. 

But as the isolated but well-publicized fatality involving a Tesla quasi-self-driving vehicle showed last May, people can put more trust in a nearly self-driving car than is warranted.  Despite warnings to keep his hands on the wheel when the self-driving feature was engaged, Joshua Brown apparently was watching a video at the wheel of his Tesla when a truck unexpectedly crossed its path, and the system failed to recognize it in time to avoid a fatal crash.  Tesla has since made changes to their system to avoid such problems, but no system is going to be 100% safe no matter how much the software is tweaked. 

What the consumers and the auto insurance industry are waiting for is evidence that over time, truly self-driving cars that require nothing more from the passenger than to sit there and not mess with things, will lead to fewer injuries and deaths than would result if all those people were driving instead of sitting on their hands.  Despite all the self-driving car test drives and public demonstrations of the last few years, we are nowhere near the point at which a reasonably robust statistical study of this type can be made.  And until that time, neither insurers nor the general public will get interested in self-driving cars in a major way.

On the other hand, fleets owned by a single entity and driving in a specific well-mapped area can make real headway, and probably will unless entrenched interests stop them, as existing cab companies are trying to do with unlicensed services. 

           

The current situation reminds me of a scene I saw recently in a 2003 movie made mostly in Germany.  Some bicyclists come to a railroad crossing with a gate lowered across it.  Now in the U. S., railroad crossings with gates are completely automatic—some track-sensor gizmo lowers the gates when a train passes by and raises them afterwards.  But in this scene, a young man in an elevated booth next to the tracks finally looks up from the book of poetry he's reading and walks over to a crank and turns it by hand to raise the gate. 

There in a nutshell you have the two choices we face regarding self-driving vehicles.  I don't know what combination of union rules and tradition and exaggerated concerns for safety led to preserving the job of crossing-guard keeper in Germany some eighty years after the technology to eliminate that job became available.  But if in 2060, we still have medallioned cabs in Manhattan manually driven by immigrants who can't find a better job and 40,000 traffic deaths a year in the U. S., it won't be because the technology isn't available.  It will be because human organizations and political factors intervened to stifle the change for fifty years.  And if for no other reason than for the sake of those whose lives will be lost to automobile accidents in that time, that would be a shame.

Sources:  Tom Votek's article "At the Los Angeles Auto Show, Industry Ponders Its Digital Future" appeared on Nov. 17, 2016 at http://www.nytimes.com/2016/11/18/automobiles/autoshow/los-angeles-auto-show-digital-future-of-industry.html.  Dr. Hans's blog appears at http://www.driverless-future.com/ and is sponsored by Inventivio GmbH of Germany.  A report on the commercial driverless-car taxi service in Singapore appeared at http://bigstory.ap.org/article/615568b7668b452bbc8d2e2f3e5148e6/worlds-first-self-driving-taxis-debut-singapore.  The movie in which the hand-cranked crossing gate appeared is "Schultze Gets The Blues" released in 2003 and written and directed by Michael Schorr.