Autonomous Vehicles, Where Are You?

“The African region has 2 percent of the world’s registered vehicles but a disproportionate 16 percent of the world’s road traffic deaths,” said Tami Toroyan, a technical officer in the department of violence and injury prevention at the World Health Organization in Geneva.

(Photo: Nissan)

(Photo: Nissan)

We’ve talked in the past about the massive potential upsides of self-driving vehicles. Just this week came word that Nissan hopes to bring autonomous vehicles to the market by 2020. If you read this heartbreaking Times article by Nicholas Kulish about a series of bus crashes in Kenya (from which the quote above is taken), you may be ready for such vehicles even sooner.

There are of course many barriers to get past before the world is ready for autonomous vehicles — yes, there will be lawsuits of all kinds and yes, professional drivers all over will protest the loss of jobs and yes, there will be people who trust a human driver more than a computer driver — but I do wouldn’t be shocked if my grandchildren grow up in a world where “driving a car” seems like something that cavemen used to do.


It seems like the driver-less system would only function if 100% of the fleet is autonomous. It seems unlikely that they would ever build a car that doesn't have a manual override feature built-in.


Have you ever used a GPS?

If you have, you would know how often they are WRONG! The re-set on them is an absolute NECESSITY. Without GPSs, driver-less cars would be useless; with the current inaccuracy of GPSs, driver-less cars would be unreliable and USELESS.

GPSs would have to be MASSIVELY IMPROVED to make driver-less cars useful, let alone safe.


Even currently existing driverless cars don't rely on GPS for any immediate decisions - they actually react to traffic signs and the movement of cars around them.


I was not talking about "immediate decisions" which GPSs obviously do NOT deal with. GPSs deal with DESTINATIONS - going from A to B. Or will driverless vehicles determine the destination for the driver?

Joe Dokes

There are a number of reasons for the high death toll in Africa, few of which would be mitigated by autonomous cars.

1. Poor infrastructure, in the west traffic engineers have spent decades and billions of dollars building and more importantly re-building highways to make them safer. An autonomous car is only going to be as good as the road it drives upon.

2. Poor quality of rolling stock. Cars in the west have an average lifespan of 11 yrs. Much of the rolling stock in Africa is significantly older than that. In addition it is in poor condition. It is doubtful that Africa could afford the cost of self driving cars.

3. Lack of clear and consistent enforcement of traffic laws. Do they even have traffic enforcement? One of the main reasons that Google cars can safely navigate in the US and Europe is that the computer can assume the behavior of other cars on the road. One merely needs to watch the dash cams of many foreign countries to see that driver flagrantly violent common driving norms.


Joe Dokes



1. I don't see how bad road quality would affect driverless cars *more* than manually driven ones, though.

2. While it would certainly take some time for new cars including this technology to show up in Africa, there is no reason why they would be more expensive in the long run. I bet that 20 or even just 10 years ago you'd have said that it would be doubtful that Africa could afford the cost of cell phones. In reality, cell phones (and now smartphones) have become dirt-cheap and are revolutionizing Africa's access to communication and information.

3. A driverless car can react much faster and better than a human driver to any misbehaviour of other cars.

Steve Cebalt

I believe auto-pilot cars will arrive -- in some form or fashion -- much sooner rather than later; only a lack of imagination stands in the way. Auto-pilot works fine on airplanes, when hundreds of lives are at risk in one large vehicle. And just like on planes, it will require a human overseer and may not apply to all routes -- perhaps just highways. All that's needed is the development of a critical mass of users -- a barrier that all new technologies must overcome. Just like autos in Henry Ford's era, when adequate roads didn't exist to support the mass sales of cars. The economic incentives are powerful: think of the time saved by drivers who can work (or whatever) during commutes. The reduced need for policing if cars are programmed not to exceed speed limits. The reduced insurance costs, collision-repair costs, and funeral costs from safer traffic.

Safety is a key point. Another poster mentioned that GPS is often wrong and so autonomous cars would be prone to problems. But there are so many market incentives to address the safety problems that it is only a matter of time and user acceptance. Frankly I would trust a pre-programmed car more than the humans I see on my local roads distracted by cell phones, texting, GPS systems, etc., or under the influence of drugs and alcohol. Humans seem far more prone to error than an effective machine. Besides, safety is not always the paramount issue. A buggy pulled by a horse is safer than an SUV driving 70 mph.

It is amazing how little autos have advanced since the Model T first made cars mainstream. The barriers that Stephen itemized are not insurmountable, if the economic incentives are properly aligned.



I can't wait. have a third lane on the interstate. kick back and listen to a book or nap. city traffic way too problematic.