IEEE Technology and Society Magazine - June 2016 - 41

now coming into view. According to the National Conference of State Legislatures, six states have enacted laws
permitting regulated use of autonomous vehicles on
public roads, and another sixteen states worked on such
legislation in 2015 [1]. These lawmakers are reacting to
the rapid progress of autonomous vehicle technology.
Throughout the automotive private sector, the debate is
no longer whether but when autonomous vehicles will
achieve self-driving skills superior to humans in terms of
safety and efficiency. Dr. Egil Juliussen, Research Director & Principle Analyst for HIS Automotive - a leading
market and consumer research firm - estimates that by
2025 self-driving vehicles will be emerging as an ordinary part of our daily commute, and by 2035 roughly
90% of the vehicles on the road will possess high levels
of self-driving automation [2].The consensus among
industry experts holds that autonomous vehicles will
eclipse manual driving as soundly as motor cars
eclipsed horses at the beginning of the last century.

Why is this Happening and Why Now?
The National Highway Traffic Safety Administration
(NHTSA) reports there were 32 675 people killed in U.S.
motor vehicle crashes in 2014 and 2.3 million injured [3].
These 2014 statistics are not unusually high or low, and
that explains, in part, why we are somewhat desensitized
to them. In another study, NHTSA reports that drivers are
the critical reason for crashes in 94% of cases [4]. Aside
from the horror of tens of thousands of violent deaths
annually and the tragic familial mayhem in the aftermath
of each, there are massive economic consequences.
NHTSA estimated the total cost for traffic crashes to be
$277 billion for the single year of 2010 [5]. As a reference
point, in fiscal year 2014 the United States spent $95 billion on the war on terror [6]. (Also note that the money
spent on terrorism is in response to a total of 3380
American citizen deaths over a period of 13 years [7].)
The long U.S. history with traffic accidents has dulled
awareness of their crippling effect on the U.S. economy.
Given these grim statistics, it comes as no surprise
that NHTSA is interested in the advent of autonomous
vehicles and their potential to eliminate death due to
human driver error. In May of 2013 the agency published a policy on automated vehicles classifying five
levels of vehicle automation, and providing states in
the U.S. with preliminary guidance for testing the various levels on public roads [8]. That 2013 NHTSA document enthusiastica lly cites anticipated sa fet y,
environmental, traffic efficiency, and economic benefits of automated vehicles.
Why has NHTSA only recently been spurred to publish an autonomous vehicle policy? The answer has to
do with a confluence of technologic advances in computing power, software development techniques, and
june 2016

∕

advanced sensor designs, as well as a parallel series of
stepping stone successes in government and industrysponsored robotic vehicle competitions. These, in turn,
have driven commercial successes in the introduction
of modest driver assist technologies that are incrementally approaching full automation. This assemblage of
technological advances and incremental successes has
triggered a sort of collective dawning in the minds of
regulators, automotive industry veterans, and - just as
important - Silicon Valley types. That collective dawning goes something like this: "Fully autonomous driving
really can be achieved. It will be achieved soon, and it
will be big."

A large swath of human choices
regarding operation of vehicles,
response to driving hazards, and
compliance with laws will be
consigned to computer software.

A long list of traditional vehicle manufacturers including GM, Ford, Toyota, BMW, Mercedes, VW, Audi, Volvo,
and Honda are investing heavily in the development of
self-driving cars. More surprising, non-traditional, West
Coast U.S. players like Google and Uber are betting big
on not only participating in the autonomous driving
arena, but on shaping the playing field. As of this writing,
Google's experimental autonomous cars have self-driven
more than 1.2 million miles - a statistic they proudly
cite alongside video links to non-actor elderly and handicapped folks joyfully experiencing the sort of independence a test ride in Google's driverless car promises [9].
Some industry pundits speculate that Google is positioning itself to become to the auto industry what Microsoft
is to the computer industry - the prime purveyor of operating systems and software applications. Another nonautomotive company, Uber, intends to both lead as well
as to "ease" the transition to self-driving cars by investing in their own R&D program and morphing their ridesha re business into a d r iverles s on- dema nd
transportation service [10]. Though no Uber self-driving
prototypes have been unveiled just yet, the company's
level of seriousness about this market is evidenced by
Uber CTO, Thaun Pham, publicly communicating his
2015 hiring goal of 1000 engineers [11]. Not to be outdone, Uber's fiercest competitor, Lyft, has partnered

IEEE TEchnology and SocIETy MagazInE

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - June 2016