Maximum number of cars added to compare list.

What's your postcode?

We need your postcode in order to provide accurate search results.

Enquire

Enter your first name
Enter your last name
Enter your phone number

Got a part exchange?

Tell us your reg plate and receive a part exchange valuation on your car?

What's this?

Compare cars side by side to save time clicking backwards and forwards between them.

Google’s driverless car is no match for the snowBack

DriverlessCarDespite the promise of Google’s driverless car revolutionising the way we drive in the UK, Google has admitted that, at present, it is no match for the snow.

Self-driving car project director Chris Urmson recently admitted snow is a struggle for the car. At the Detroit Motor Show last week, he suggested that Google should test its car in more challenging situations as opposed to its current testing ground in California where it “doesn’t snow”.

Aaron Brindle, a Google spokesperson said: “When it’s snowing or really foggy, a human driver has limited visibility — so too do our vehicles.

“Though we’re working on improving this, the good news is that, in the meantime, our cars recognize when they have limited visibility and will make the safe decision not to drive.”

This latest observation has done little to encourage experts who believe that Google is still decades away from a solution. They believe that obstacle detection technology has a long way to go before it can cope with the winter elements.

Some of the problems experts have identified relate to the driverless car’s computer vision systems which rely on the roadway ahead to understand where the lane boundaries are – this is impossible if they are covered with snow. Furthermore, experts have cited that light-detection sensors used by driverless cars can be tricked by the snow into thinking there is a “phantom obstacle” in the way.

131212 Icy roadsThe same problems occur in icy conditions, where experts believe that computer visions are incapable at present of sensing black ice even when equipped with built-in digital maps, GPS technology and laser scanners for positioning and obstacle detection.

When a self-driving car cannot recognise the snow and ice and starts to feel challenged, it will either stop or return control to the human driver, similarly the car is programmed to stop when it encounters another car backing out of a driveway, a construction zone, an animal scampering across the street or debris left on a roadway.

Taking the above into account, experts have admitted it will be decades before automated cars will be capable of dealing with unexpected elements.

Massachusetts Institute of Technology mechanical engineering professor John Leonard said: “It’s important to not underestimate how hard the remaining problems are. We shouldn’t get our hopes up too much that it is all going to happen overnight.”

Neil Addley, managing director of Trusted Dealers added: ‘There’s no doubt that the developments in driverless technology are impressive. However, as motorists who have seen recent road conditions will be all too aware, its very hard to predict exactly how those conditions will change as a result of snow and ice. And so it’s not surprising that driverless cars aren’t able to adapt as quickly as humans. Simply electing not to drive seems like a bit of a copout.’

 

 

Posted by Leana Kell on 21/01/2015