GarryP

See the difference?

If you look at the operating cost structure of a aircraft, you'll find that pilots are a relatively small component. Obviously it depends on the size of the aircraft, but you can assume that the pilots are less than 10% of operating costs - they don't get paid very well. Not the same scope for savings.

Even so, chatting with a pilot mate of mine, it looks very like single pilot operation will push up through the size categories in the foreseeable, and obviously particularly for cargo aircraft, entirely autonomous operation is likely eventually - just big drones. Part of the problem is the failure of the likes of CASA to equip for full cat 3c operation in Australia, but advances in localisers are likely to make that moot. It's likely to be issues of flexibility that push fully autonomous flight operations - probably once autonomous ground vehicles are accepted by passengers.

GarryP

Ford's self-driving car 'coming in 2021' - BBC News

Here we have Ford saying yesterday they are going to deliver robotaxis, without steering wheels, by 2021.

At the moment it's pretty much a question of which car or IT company is NOT targeting autonomous vehicles and 2020.

Beoz

It can all be done now. Just gotta get past the legislation and will incidents like this, 2020 or 2021 is unlikely.

US opens investigation into Tesla after fatal crash - http://www.bbc.co.uk/news/technology-36680043

Reading your thread it uses Lidar to figure out what's around it. Lidar is very reactionary rather than proactionary. Proactive driving thought is needed. Something only a human can do at this point.

GarryP

That doesn't make a lot of sense. Lidar is just a sensor, how you use it is up to the software, and the AI processing is coming on in leaps and bounds. Hell the Lidar based DARPA vehicles were completing cross country courses ten years ago.

The sense making from scene sensor data is probably already ahead of humans, if for no other reason than it doesn't get bored.

ozzieeagle

About time we had some legislation news re driverless in Australia.

Beoz

Of course it makes sense. Lidar can't predict the future or guess what a driver might do. A driver can anticipate what another driver might do through experience.

GarryP

But an AI can, that's the point after all. It's nothing to do with sensors

The only instance where the Google car had issues was when it expected the bus to allow it room at a junction that it didn't. It was predicting what the bus driver would do.

Beoz

And there you have it. There's a long way to go in automated predicted behaviour.

GarryP

Err nope.

The point is how well it has been doing. The performance has been pretty safe - perfection is not the benchmark, an asian granny with two screaming kids in the back is.

GarryP

Airbus wants to build flying taxis because everyone hates traffic - Autoblog

So, no sooner will there be autonomous cars, than the skies will be full of autonomous helicopters AND delivery drones.

Beoz

Its already happen. Has been for years with a pilot.

http://onemileatatime.boardingarea.com/2015/01/21/gotham-airs-new-99-helicopter-manhattan-jfkewr/

Amazulu

ozzieeagle

Uber to start first driverless fleet in Pittsburgh this month.

Uber’s First Self-Driving Fleet Arrives in Pittsburgh This Month - Bloomberg

Otto apparently are looking at driverless trucks across the Nullabour.

Otto: Inside Uber's New Self-Driving Truck Division

GarryP

Yep, the most interesting bit was :

Now Otto are 'the thing' in truck automation - and Uber in trucking would be like putting the Sabre Tooth Tiger amongst the pigeons.

People keep saying "oh, it won't happen in my lifetime" - ignoring that they are already putting pilot projects on the street in 2016. Something like automated trucks in virtual roadtrains across the Nullabor is something of a nobrainer. Great spot to test it out in fact, prior to freeways in the US.

Beoz

Anyone got one of those driverless vacuum cleaners? Any goo Something to make Mrs Beoz redundant.