In a few of our past posts, we have mentioned that the introduction of fully self-driving cars is not a matter of “if” but more or less a question of “when.” Indeed, some automakers have noted that they will offer such cars by the end of the decade. However, it may not be so easy when considering the possibility of technical problems causing accidents.

This conundrum is ironic considering that the main focus behind autonomous cars is so that the potential for accidents could be reduced. After all, a self-driving car would not get too tired to drive and drift into oncoming traffic. It would not drive while impaired by alcohol, or would not drive aggressively (or while distracted by something else). Nevertheless, there have been several instances suggesting that autonomous cars would still be vulnerable because of technical inconsistencies that may not be attributed to human error. 

Essentially, humans carry the capacity to bend (or even break) traffic rules in some situations in order to avoid an accident. Computerized cars would not necessarily have this capacity, and this could lead to crashes in situations where human drivers could avoid them. Indeed, one of Google’s testing engineers insisted that many of the accidents their vehicles have been in were when the car has been stationary, and neither the computer nor a human was driving.

Regardless of the trust or ambivalence that you may have in driverless cars, there is certainly a market for them, and they will be coming sooner than we think.