Shard

It seems ambitious to me, but I'm not in the AV industry so it's hard to judge. Once the tech is ready (and it appears to be at the cusp) it's only regulatory approval that will slow it down. Perhaps taxi services will be limited to zones or specific routes initially. For example a robotaxi from LHR along the M4 into Chiswick does not seem so incomprehensible.

GarryP

Well Google already have cars driving around as taxis, without anyone in the driver seat. And Tesla seem to be firming up on delivering autodrive (and new faster hardware to service it). Ford spun off the autonomous division (they might well be looking to leave the rump of the company with all the debts and liabilities).

So my guess is that another three years and someone will have it working.

As for regulation, if nobody else will move fast, the chinese will.

Shard

Yes, a lot can happen in three years. There's another article on Addison Lee/Oxbotica in The Guardian which suggests that they will go gangbusters to set it up. Good point about China, once they have an AV system set up I think the US will become far more focused on deployment if only to save technological face.

Amazulu

Legislation will be a massive handbrake - plus moral and ethical issues. If the technical and legal areas are not 100% assured then it's a no go

Unless a driverless taxi is going to take you directly and exactly where you want to go, then what's the point?

I'm sticking with at least 2030

Beoz

I agree. Most of the planet is still stuck on train drivers and all they have to do is speed up and slow down. If governments can't see the benefit in fitting out train fleets with driverless then a listed trucking company has no chance. Shareholders will just say "stick with drivers - I want a return this year, not in 20 years"

GarryP

Truck companies will change both because once you have solved the issue for cars, trucks aren't vastly different, and because you are talking a good 25%+ in driver related costs (wages, breaks, etc.).

So, you are looking at retail jobs, food jobs, call centre jobs, driving jobs, office paperwork pushing jobs, and of course, manufacturing jobs. You can readily see where 50% unemployment comes in.

old.sparkles

On autonomous vehicles, this is an interesting read - https://www.ri.cmu.edu/pub_files/pub...her_2008_1.pdf

The challenge was 2006 and the vehicles had to navigate surfaced roads, rough roads, car parks, traffic, etc. I think the speed averaged was only 48km/h but given the advances since I think we could see autonomous vehicles available before 2030.

Shard

​​​​​​Admittedly it's odd that rail employees still wield bargaining power for their jobs. This week some rail drivers in the UK are striking because of plans to do away with the guards. However, I don't think a direct comparison with trucking can be made. The cost of the guard/driver in relation to high fixed capital costs and high revenues from passengers is minimal; whereas the cost of a driver for one truck and it's cargo is relatively high. Human safety (perceived or real) also impacts the rail scenario.

Once AV tech is available for trucks, it will be implemented without delay, and truckers automated out of existence. Maybe they can start opening up wine bars.

Amazulu

Driverless trains have been around since the 1970s (BART) yet most trains today still have drivers - despite the fact that operating trains without them is relatively easy

Rio Tinto (an international leader in the technology) is in the process of taking their haul trains in the Pilbara driverless. I was at uni with one of their guys who is integral in this project. He was telling me that going driverless is not necessarily cheaper (they need to have people on standby 24/7 in case anything goes wrong) but it's rather a big f**k you to the CFMEU

Beoz

Exactly. People are cheap. Driverless will always be about profit and productivity.

Speaking of which, in the back of a cab trying to avoid a union march in Melbourne. These guys have no clue. If you help business make money you can demand one for your members. Without the give, there is no take. Victoria should have a number plate swapsie with Queensland "The Smart State"

Beoz

So explain how in this human safety concerned world do we have some driverless passenger trains in some places and not others?

I will tell you. The cost of replacing a human with a machine does not provide a return on investment.

Same with trucking. You aren't replacing one, you are replacing many. Slow roll out if and only if the shareholders are willing. And that's even after the legislation gets out of the way. 2035 at best.

ozzieeagle

Labour shortage here now

https://www.abc.net.au/news/2018-07-...onomy/10050678

Hopefully, this Suburban rail loop in the planning stage will be driverless, that way there will be more trains and worth travelling outwards from the city for.

Lots of infrastructure being built in Victoria at present and more to come.

Shard

My point was that the economics of drivers on trains differs from drivers on trucks. Driverless passenger trains depends on the location and type of train.

Wol

Good luck with Chinese autodrive...

There's an interesting article in NS last week about the huge energy requirements of our "paperless" world (yes, that's sarcastic...) what with vast amounts of electricity being used for blockchains - and, surprisingly, for the AI used in such things as autonomous cars.

Regarding Tesla (and I'm a big fan, trying to get Mrs Wol to allow me one) ten minutes on any of their user forums would convince anyone that they are nowhere near going for full autonomy. The number of near misses that are reported is staggering, when you consider the hype.

Wol

That's exactly the point: unless they can go literally anywhere they are useless. With a "safety" driver, OK (perhaps) but then what's the point?