By Dominik Natter
With almost 70 million cars produced worldwide each year, even small savings per vehicle can have a
great impact. This could mean a reduction in air resistance leading to lower fuel consumption, or an
improvement in catalytic converters reducing emissions, or simply a replacement of the driver for
an…. Wait, what? Yes. There are already functioning systems capable of driving all on their own, leaving the driver sitting there both useless and speechless. Actually the first driver less car presented was
the Linrrican Wonder in 1925; however, it was radio-controlled. Still, experts regard trials done already in the 1950s as the first promising experiments with autonomous cars, using wires embedded
in the road for orientation. Today’s technologies are far more reliable, and are equipped with a highly
sophisticated network of sensors including LIDAR systems, cameras, gyroscopes or radar. Autonomous cars may come in a variety of forms, including the possibility of manual driving by activating a
steering wheel on request, which is a service many driving enthusiasts may take advantage of.
The completely autonomous option offers great potential. Most importantly, the number of road
fatalities could be drastically reduced. The system is able to avoid the most common causes like
speeding, disregarding priority, or drunken driving, and is more likely to adjust appropriately to external factors like bad weather conditions or changes in road surface. Above all, this will save lives,
but also cut costs.
Cost saving will also be achieved due to more efficient driving (which may lower fuel consumption or
time spent in traffic jams) and a remarkable increase in productivity. All the despairing people
fighting their way through the usual morning and evening traffic jams to get to work or university and
home again, keep dreaming of how to spend this time in a better way. As passengers rather than
drivers, they could enjoy breakfast, call friends, prepare documents for work, or even take the first
nap of the day. With an autonomous car the possibilities are endless, and the time wasted on driving
is converted into efficient use, contributing to the country’s GDP.
Sounds good so far, doesn’t it? Yet the most promising use of autonomous cars is car sharing. Cars
are said to be parked 95% of their time. However, shared automobiles could be on the street all day,
stopping just for refueling or if not needed. Governments could use these cars for expansion of public
transport and finance them in the same way as subways or buses. Thus people wouldn’t need to bear
the high costs of an autonomous car. This could find expression in a reduction of vehicles on the road
by a factor of 10. Air pollution would be reduced and the streets less congested, leaving space for
bikes, pedestrian areas and parks. Of course fewer cars lead to fewer accidents too.
So far many of you are left shaking your head wondering why autonomous cars aren’t state of the art
yet. First of all, technology didn’t allow a breakthrough much earlier (mostly because of missing
computing power). But now, as this problem seems solved, there are still two severe problems: safety and the question of liability. If your autonomous car crashes into another autonomous car (which
won’t happen too often of course), whose fault was it and who has to pay the bills? The drivers? An
insurance company? Or one of the manufacturers? Highly detailed documentation will be needed for
a decision based on facts. Whereas this is a legitimate subject for debate, safety is not. The autonomous car software needs to be perfectly protected against hack attacks. The possibilities for hackers
for causing damage are humongous, extending from deactivated brakes to crashing full speed into a
crowd of people. Possible preventative measures include new encryption methods developed solely for the communication between these cars, so that incoming signals can be determined as safe or as
from an unknown source. Letting people abruptly take control over the car again is no possible solution, as researchers of General Motors explain that people “largely disengage from the driving task”
after a short time of enjoying an autonomously steered ride. Thus this spontaneous change to manual driving seems very unsafe. Anyhow, 100% safety can’t be guaranteed yet and probably will never
be, as hackers will always adapt and improve their methods. Thus, car makers should really focus on
the improvement of the cars armor and hang in there.
Still, the trend toward autonomous cars shouldn’t be detained; its advantages are just too significant.
Governments, companies and we the people should cooperate to push this invention to market maturity with all speed. The next step then needs to be the expansion of vehicular communication systems. Getting data from each other as well as from central/local traffic control systems makes autonomous cars even more efficient, allowing them to drive faster and avoid building sites and traffic
jams, branding road signs as obsolete. As always in robotics, the impact on employment has yet to be
determined. Increasing productivity, replacement of taxi drivers, fewer car repairs or shrinking car
production all cost jobs. Obviously new kinds of jobs will also arise, yet it would be advisable to think
of solutions for this upcoming job shift already. Governments, experts, companies and employees
need to come up with satisfying concepts for all parties. If they manage to do so, this is going to be
one hell of a ride.
4th semester student, Bachelor program “Mechatronics/Robotics”, UAS Technikum