Putting the brakes on autonomous driving

More than a million people around the world die in road accidents each year, with human error cited as the cause in 90% of the cases, the majority of which involves persons of a young adult age.

Malaysia has the unenviable distinction of being ranked top three in the world by the World Health Organisation for her road accident death rate of 23 per 100,000 persons, which equates to around 7,000 fatalities every year, or 20 individuals per day.

There is little doubt that autonomous driving technology will eventually be far superior than human beings at preserving lives on the road, but the untimely demise of Elaine Herzberg in Tempe, Arizona recently, caused by an Uber self-driving vehicle* that seemingly failed at what it was supposed to do, is a timely reminder to auto manufacturers, ride hailing companies and technology peddlers that this piece of technology should not be rushed through to the public domain.

The science behind a fully autonomous car is mightily complex. It is not to be confused with driver assistance functions such lane keeping, active braking and adaptive cruise control which are already commercially available.

While manufacturers are pushing on with the development of self-driving vehicles equipped with sophisticated laser sensors such as LiDAR (Light Detection and Ranging) along with greater computing power (read artificial intelligence) to crunch the data, we, as consumers, must be wary of the rosy picture being painted of the future.

Instead, we should pose the question that given the unimaginable variables in real life situations, can an autonomous vehicle (as currently defined based on different levels of automation) ever be smart enough to function independently and safely?

Like a modern railway system, it’s hard to imagine self-driving cars not requiring infrastructure support at some point to be truly autonomous, particularly within the confinements of a city.

Without an intelligent road system capable of tracking and communicating with cars, there is no redundancy whatsoever for other road users such as pedestrians and bicyclists to rely on; a car doing its own thing is indeed a frightening scenario.

But because auto manufacturers and lawmakers operate at vastly different speeds and motivations, a smart road system supporting self-driving vehicles is an inconvenient but critical aspect rarely touched upon.

Many lessons will be learned from the first ever death of a pedestrian inflicted by an autonomous vehicle. But unlike the more pertinent issue of hydrocarbon emission reduction (because global warming isn’t a hoax) through the proliferation of electrified vehicles, the glorification and the ‘rush’ to adopt autonomous driving technology may have to be reviewed, if not regulated.

(* The Volvo XC90 was operating on Uber’s self-driving technology during the Arizona accident.)

Denis Wong

Denis Wong

In the age of misinformation and spin, Denis prefers candour and a counterpoint, because the truth matters.
share on: