Skip to main content

Why Purely self driving cars may not work in major cities


Google's Lexus RX 450h Self-Driving Car

Self-driving car manufacturers have arranged the car so that a human driver can take over once they get to a tricky spot.  But why is this? Driverless cars don’t get road rage, drunk or sleepy. They have had a few technical hiccups, but these are being ironed out as technology improves.  According to the US Department of transportation, more than 90 percent of car accidents in the United states are caused by driver error.*
     Self-driving cars have the rules of the road immaculately programmed into them, so that makes them perfect for driving in large cities, right?
Wrong.  The times that I have spent driving in Dublin, Ireland and America and seeing the state of driving in big cities in general tells me otherwise.  It is not a flaw in the car itself.  It is just that one must be aggressive to drive in major cities.  I will never forget the day that a sleepy truck driver in America basically pushed my wife and I off the road.  Having spoken to a friend who has been involved in trucking deliveries all over Europe, he informed me that in cities like Rome and Paris, even people on pedal bikes will cut him off.  Roundabouts are the most insane part of the whole journey.
 If one gets stuck at a tricky place in a city, it could be some time before one can make a move to get clear of it.  If one gets out, it is because someone is decent enough to let one out (unlikely), or a fast one is pulled.  If one waits for the right, polite, opportunity, then one could be waiting forever.
            If this is true in major cities in America and Europe, it is every bit more so in Southeast Asian countries where traffic at junctions almost morphs together; the Tuk-tuks and mopeds nearly colliding with each other.  The ubiquitous horn is merely to let everyone else know that one intends to occupy a space in traffic, whether they like it or not.
            The rules of the road in these countries are more theoretical than practical.  Traffic lights in cities are rarely strictly obeyed. This is no less true in Africa and Latin America than Asia. 
Additional challenges to the driverless car are the plethora of human cues such as hand signals that humans recognise but machines cannot.  One also cannot ignore the large number of Darwin award nominees who run unexpectedly onto the road.  People, through experience and intuition, know when other people are about to do something stupid and dangerous.  The car’s computer must make a decision based on the input from the surrounding environment.  It must know who will try and cross a road instead of remaining still.  Pranksters will undoubtedly make use of this inability and will trick the computer into stopping the car.   The end result will be a traffic jam at best and a fatality at worst. 

Car companies will make leaps and bounds with artificial intelligence, but they should leave the controls in place, just in case.

*https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812115

Comments

Popular posts from this blog

Whatever happened to slingers?

Andean sling - Img source Wikimedia commons This is a question that crops up in forums and discussions about ancient combat.  Slingers were described by the ancients in very positive terms.  Suddenly they were no longer present.  There are scanty records of sling combat post-antiquity and many explanations have been postulated for their demise. The first is that slings were replaced by bows.   Bows became predominant simply because they were better, more precise, farther reaching etc. The problem with this theory is that while both slings and bows are biodegradable, the oldest discovered bows (Holmegaarde, 6000 BC) predate the oldest sling discovered (2500 BC, Lovelock cave) by thousands of years.   The theory also discounts all written evidence concerning the sling.   Slings were crude but powerful weapons which could project missiles much further than an archer could shoot an arrow.   Long range throwing also depended on the type of ammun...

Vikings vs Romans: A hypothetical battle

Roman army vs the Vikings If one were to take the Roman army at its height, it would decimate any early medieval army that would cross its path, according to a claim by Dan Carlin. To test his theory, let’s take a look at a hypothetical battle fought between the Vikings and the Romans.   It is easy to see how Carlin would come to such a conclusion.  At the height of the Roman empire in 117 AD, the army boasted hundreds of thousands of soldiers, all professional, all equipped and supplied well at the expense of the state. No such force existed in western European countries in the Viking age.  Armies were mostly levies with core body of professionals such as the Germanic huskarls or household guard. One could argue that this lack of professionalism was mitigated by a warrior culture.  This is erroneous, however.  Classical Greek and early Roman societies had a warrior culture that prized courage in battle.  Every citizen was also considered a so...

Does Gandhi's civil resistance always work?

Gandhi's salt march Non-violent civil resistance is a wonderful idea whose utility cannot be underestimated.  In theory and practice, the authority which does not lead by consent will naturally lose its foundation.  If the non-violent civil resisters are brutalised , then the authority loses all vestiges of legitimacy since legitimacy stems from honorable action and popular consent. However, it is not a principle that can work universally.  Non-violent civil resistance worked well in India for a number of reasons. The first is the population in India outnumbered the British expatriate class and administration by over a thousand to one.  Secondly, the vast geographical expanse between India and England was a further deterrent to keeping India.  Keeping India could only be justified (from an economic perspective) if the risks and losses from such a distance could be mitigated or exceeded by the resources gained from ruling.  Furthermore, the adminis...