It seems to me….
“Self-driving cars are the natural extension of active safety and obviously something we should do.” ~ Elon Musk.
The rise of the automobile in the early 20th century transformed many aspects of modern society: where we live, what we buy, how we work. As cars and trucks became commonplace, they created whole classes of jobs and made other professions obsolete. We are now on the cusp of an equally transformative technological shift in transportation: from vehicles driven by humans to vehicles that drive themselves. Far more profound transformations will follow once cars and trucks can be trusted to pilot themselves routinely – even with no one inside. The long-term impact of autonomous vehicles on society is hard to predict but also hard to overstate. The only certainty is that whenever this technology becomes ubiquitous, life will be different than it had been before.
There is plenty of room for improvement on that front. Auto accidents currently kill 1.2 million people every year, more than 30,000 die and some 2.3 million are injured in the U.S. alone. It is obvious that humans are too dangerous to be allowed to drive. Self-driving systems may have bugs – the software that runs them is complicated – but they are free from the myriad distractions and risk-taking behaviors that are the most common causes of crashes today. In the near term, semi-autonomous safety systems that engage only to prevent accidents, but that otherwise leave the driver in charge, will also likely reduce the human cost of driving significantly.
All major automotive companies are working on the development of autonomous self-driving vehicles. Google’s vehicles have driven over a million miles with only one minor vehicle accident possibly attributable to their vehicle (seventeen other accidents were the fault of human-driven accidents).
Once autonomous vehicles become readily available, probably in about five years, how long will it take for human-operated vehicles to no longer be permitted to be driven? At some point, most people will prefer to not even own a passenger-type vehicle similar to an increasing number of city-dwellers. Exclusive car ownership could then cease to be the necessity of modern living that it is today for so many people. Shared cars and driverless taxi and delivery services could become the norm.
Driverless vehicles will be able to outperform humans on routine driving tasks but it will be quite some time before they are able to react to rare events. Driverless vehicles have greater perceptive abilities, faster reaction times, and are not distracted or become drowsy, hungry, or troubled by bodily functions. They still have difficulty adequately performing in inclement weather condition, reacting to human-directed traffic, or on roads that have not been thoroughly mapped. It is highly unlikely that all possible situations can be preprogrammed such as ethical dilemmas; e.g., choosing between two undesirable alternatives such as striking a pedestrian or hitting a pole or another vehicle.
According to the U.S. Department of Transportation (DOT), there are 4.5 million automotive vehicle accidents every year. Additionally, 4.2 billion hours and 2.8 billion gallons of fuel are wasted in traffic congestion annually. Though quite a few companies are developing fully autonomous self-driving vehicles, they are not expected to be publicly available for several years.
The National Highway Traffic Safety Association (NHTSA) has established a description of automation levels describing the degree of vehicle control for driver/vehicle computerization:
Level 0 (manual): Basic vehicle operation devoid of automated system support.
Level 1 (function specific): Automated operation for single control normally invisible to driver (antilock braking, brake assist, electronic stability control, electronic traction control…).
Level 2 (combined functionality): Simultaneous automated control of two or more functions; e.g., highway pilot or traffic jam pilot; that combine adaptive cruise control capable of adjusting the targeted cruise speed depending upon a preceding vehicle’s increase/decrease in speed combined with an automated lane-control system.
Level 3 (limited self-driving): Fully automated control of all driving related tasks for extended periods. Human intervention (conditional driver takeover) still necessary for unusual or unexpected situations exceeding ability of automated system.
Level 4 (full automation): Fully automated control of all driving tasks beyond selection of destination and emergency stop button.
Prior to driver-less vehicle general availability, other technologies are being developed that are intended to increase vehicle safety and efficiency. In addition to 2016 requirements for backup cameras, additional technologies identified by the DOT as Vehicle-to-Vehicle (V2V) and vehicle-to-infrastructure (V2I) or collectively as V2X. These initially include “Intersection Movement Assist” warning of intersection cross-traffic, “Left Turn Assist” warning of approaching traffic from opposite direction when making a left turn, collision avoidance, parking assistance, and several other technologies. V2X technologies extend beyond a single vehicle providing capabilities unable to be provided in single-vehicle systems such as warning of conditions normally undetectable by sight or radar or at distances beyond 300 meters.
Automakers are reluctant to support V2X technologies without agreement on communications standards upon which these features are dependent. The CAR‑2‑CAR Communications Consortium in Europe and the Intelligent Transportation Society of America are developing such standards.
Transitioning from human-driven to autonomous vehicles will initially encounter difficulty with subtle vehicle-driver expectations when autonomous and human-driven vehicles share the same road. Visual recognition between drivers at four-way stops; gestures/nodes between pedestrians and vehicle drivers…. Autonomous vehicles will encounter legal, ethical, social, and technical challenges with the technology possibly being the easiest to solve. Expectations will necessarily have to change.
Self-driving autonomous vehicles are becoming increasingly common on U.S. roads and highways. Their advantages in safety and efficiency brings into question how long human-driven vehicles will be permitted to remain on our roads. The era of human-driven vehicles, along with petroleum-fueled vehicles, is rapidly coming to an end.
In many ways, eliminating the human driver is attributable to the proliferation of technological distractions available in today’s vehicles. There has been a 28 percent increase in auto-related fatalities and auto-related accidents are the primary cause of deaths for those between the ages of 15 – 24, the age group most susceptible to technological distraction. Regardless of cause, 32, 675 Americans died in automotive accidents just in 2014.
All too soon, people will romantically reminisce about those good old days and the open road that stretched out endlessly before them. American’s romantic attachment with their automobile has always been somewhat flawed in that it persistently overlooked the expense, wasted time, and death and maiming it caused.
The safety record of those autonomous vehicles already on the road demonstrates not only their improved safety but also their obvious superiority to human equivalents. 94 percent of all road accidents result from driver error at an estimated annual cost of $836 billion. Transportation-related expenses, primarily automotive, accounts for about 17 percent of the average household annual budget while the average U.S. citizen spends about 42 hours/year stuck in traffic (82 hours/year in congested areas).
While initial accidents can be expected while autonomous control systems become more fully developed, anticipated improvements in vehicle safety mandate adoption of self-driving transportation as early as feasible. Everyone should look forward to the freedom of leaving the driving to someone – or something – else.
That’s what I think, what about you?
 Elon Reeve Musk is a South African-born Canadian-American business magnate, engineer, and inventor.
 Kirkpatrick, Keith. The Moral Challenges of Driverless Cars, Communications of the ACM, August 2015, pp19-20.
 Frequency of Target Crashes for IntelliDrive Safety Systems (DOT HS 811 381), United States Department of Transportation, http://www.nhtsa.gov/DOT/NHTSA/NVS/Crash%20Avoidance/Technical%20Publications/2010/811381.pdf, October 2010.
 Geller, Tom. Car Talk, Communications of the ACM, March 2015, pp16-18.
 Steinmetz, Kathy. Why You Shouldn’t Be Allowed To Drive, Time, 7 March 2016, pp 52-63.
 National Highway Traffic Safety Administration (NHTSA).