Self driving car kills pedestrian.

[Mike Sales]

Just heard that a woman pedestrian in Arizona has been killed by an Uber robot car.

https://www.theguardian.com/technology/2018/mar/19/uber-self-driving-car-kills-woman-arizona-tempe

[yakdiver]

Sadly the first on many, machines can not work by themselves they need a minder

[Bonefishblues]

Sadly the first on many, machines can not work by themselves they need a minder

It had one.

[Mike Sales]

There seems some confusion here. News media are calling the woman a pedestrian and a cyclist. It is not specified whether she was pushing the bike or riding it, but agreed that she was crossing the road outside a crosswalk.

[thirdcrank]

This reminded me of William Huskisson, a British politician who made it into the history books as one of the first - perhaps the first - people killed by a railway engine, Stephenson's Rocket.

The death and funeral of Huskisson caused the opening of the railway to be widely reported, and people around the world became aware for the first time that cheap and rapid long-distance transport was now possible – and could be dangerous.

https://en.wikipedia.org/wiki/William_Huskisson

[Cyril Haearn]

Bridget Driscoll was the first person killed by a driver

The judge earnestly hoped she would be the last

[Mick F]

Am I right in saying, that there we more people killed on the roads per year before we had motor vehicles and indeed bicycles?

[pete75]

If you follow the links through from that article it shows Uber self drive cars have been banned in California for stuff like running red lights and failing to abide by state regulations including operating self driving vehicles without a permit to do the same. Apparently Uber think their rejection of government regulations is an important matter of principle.

https://www.theguardian.com/technology/ ... al-tactics

[Stevek76]

Yes uber have had a very poor attitude to this so it's not exactly surprising that the first fatality should involve one of their cars.

[andrew_s]

Am I right in saying, that there we more people killed on the roads per year before we had motor vehicles and indeed bicycles?

Numbers are hard to come by, as official recording of road traffic casualties didn't start until 1926, but it seems that deaths were somewhere around 2000-2500 p.a. during the 1900-1910 period, before cars were widespread. That's about the same as in 2007, more than now, but about 1/3 of the mid 1930s

[thirdcrank]

Am I right in saying, that there we more people killed on the roads per year before we had motor vehicles and indeed bicycles?

I suspect that's more likely to be before the Belisha beacon man introduced the driving test. 1934?

[kwackers]

Sadly the first on many, machines can not work by themselves they need a minder

Per mile travelled they're massively safer than monkey operated cars - and more importantly are getting more rather than less safe. Every time one makes a mistake they ALL learn - imagine how brilliant that would be if applied to us!

Having said that I think there's too much of a rush towards self-driving and Uber is far more guilty of this than pretty much any other operator since they're more than a little keen on replacing their drivers.
Personally everything I see about Uber shouts "cowboy" at me. They should lose their self-driving license and let folk who develop these things responsibly carry on.

[Cunobelin]

Sadly the first on many, machines can not work by themselves they need a minder

Alas so do many Human drivers!

[pwa]

Trouble is, software seldom works 100% and you don't find the flaws until you use it. Which in this case is controlling cars in public spaces. Carrying out R&D in a public space is bound to be a risk. Anyway, they now have a bit more data to chew over and some compensation to pay out. Probably already budgeted for.

[thirdcrank]

If this is going to be just another driverless-car thread, here goes.

The technology is capable of almost anything. We are now in the early stages of marketing and one of the issues is not what the technology can do, but what the manufacturers can get away with not doing. eg To what extent are driverless cars expected to anticipate or provide for the unpredictability of human beings. The greater the safety in the presence of traditional traffic, including cyclists and pedestrians, the slower the driverless vehicle. I'll predict a lot more infrastructure intended to keep cyclists and pedestrians "out of the road."

Beyond that, I don't think the marketing men have settled on what set up will sell best. It might be a totally driverless Uber-type arrangement where you summoned a vehicle for your journey. One drawback from the manufacturers' POV would be that the logic would be more efficient use of vehicles so a lot fewer sold.

That sales model would also risk an image of users who couldn't drive such as the elderly, That would detract from the big moneymaker, a super version of all the currently available driver-assist stuff but which still left ultimate control to the "driver's judgment." Drivers increasingly depending on the system to avoid crashing but able to use those benefits to bash on regardless. Cars automatically complying with speed limits will not sell well.