5 ms·
Self driving cars are more than a "few years" out. All fully automated proofs of concept so far operate in light traffic with clear skies and high quality roads
by falcolas 10y ago
Self driving cars are more than a "few years" out. All fully automated proofs of concept so far operate in light traffic with clear skies and high quality roads. These conditions are not prevalent in most of the US.
I believe that even truckers will be safe well into the next decade. They don't add all that much cost to the overall costs of shipping goods (when compared to fuel, maintenance, and insurance).
- noir-york 10y agoI hope you are right, but the progress with deep learning and AI has snowballed in the last few years. At least that's my perception. > I believe that even truckers will be safe well into the next decade. They don't add all that much cost to the overall costs of shipping goods I do not know about the US, but in Europe labour costs are high and there are laws on number of hours that truckers can work. Self-driving trucks can be on the road 24 hours a day all year round, with no unions or human resources to manage.
- throwawayReply 10y agoNo human resources? Who's gonna fill up the tank? Fewer human resources, certainly. But humans are very flexible. In the containerised shipping world that doesn't add huge value, but for a taxi some flexibility is a value-add, be it changing your mind about your destination or falling ill suddenly (when you probably can't just log in and change your destination) are examples of when people are well served by human operators. Self automation has to become cheap enough that the premium of having people drive is worth losing compared to the price difference. It isn't enough to just have tje technology at a level it can serve.
- sokoloff 10y agoFewer people would likely die from falling ill suddenly and undetected than currently perish from human-caused accidents. I suspect fewer would perish from sudden medical events in cars overall as a medical event no longer means a likely crash.
- lukasm 10y agoSelf-driving trucks will lower the mortality rate and the insurance premiums as well as increase the utilisation of assets - trucks and storage.
- ethanbond 10y agoFunctional self-driving trucks* That's the crux of the issue. People need to stop acting like we have self-driving cars. We don't. Once we do (a few years out, at least), THEN our governments will start figuring out how to deal with it. And by "governments," the plural is important. Remember that every single jurisdiction that a driverless car goes through must approve of them. And for long haul truckers? That's hundreds of jurisdictions for any given route.
- falcolas 10y agoI don't think the insurance will go down all that much; the cost of a lost $50,000 load of bananas and $150,000 worth of truck/trailer is not going to change because of even a 50% reduction in crashes. Plus, why would insurance companies actively work to lower their profits? ;)
- maxerickson 10y agoThis random article puts the driver at $0.36 out of $1.38: http://www.findatruckingjob.com/trucking-info/trucking-articles/real-cost-operating-truck http://www.findatruckingjob.com/trucking-info/trucking-artic... That's pretty substantial. If you have 2x the utilization on a cheaper to operate vehicle, that's where new investment is going to go. Or take a retrofit kit. It can cost ~$80,000 and be a pretty obvious investment to make.
- falcolas 10y agoInteresting. I've researched this in the past and found that this number per mile varies between 5¢ and 30¢ per mile, and that it's usually charged out at $2.50 a mile split between the owners of the load. Even at your 36¢ per mile, that only amounts to a savings of $150 on a $1,000 bill to move a truck of goods 400 miles. Given that the value of a truck full of bananas is around $50,000, that $150 savings is a rounding error (especially since the AI won't be free).
- maxerickson 10y agoThe $50,000 doesn't have much to do with it, people looking to replace/expand trucking capacity will look at the $850 and $1000 and only choose the $1000 if there is some other factor that is similarly significant. Even if existing shippers are conservative, someone will definitely try to use a 10% advantage. It's also the case that the shipper is looking at how much they can make by moving the bananas, which is probably less than the wholesale cost of the bananas. So the $150 is more significant to them than the $50,000 suggests. These ads are saying $0.50 a mile for moderate experience: http://www.indeed.com/q-Cent-Per-Mile-Otr-Driver-jobs.html http://www.indeed.com/q-Cent-Per-Mile-Otr-Driver-jobs.html I think between variable fuel costs and the up and down market for shipping that a given article won't be very reliable for today, but something like "a substantial portion of $1-$3 per mile" is good enough for discussion.
- IanCal 10y agoThat 36 cents also seems to be a direct salary cost, there would be other overheads to employment. And I note that the biggest cost is diesel. I wonder if self driving trucks would be more fuel efficient?
- djaychela 10y agoNot sure about the US, but in the UK and EU drivers have a limited number of hours they can drive per day. When the clock runs out, they have to stop. That won't be the case with self driving vehicles as they won't get tired. I would think this alone would revolutionize the logistics industry, without any other factors.
- falcolas 10y agoThe US also has limits, but you can go a long ways in 11 hours. And when a load really needs to get moved, you can get up to 22 hours with tandem drivers. The cost of drivers in relation to the value of the cargo being moved is usually a rounding error.