• The dispute between MobilEye and Tesla regarding the appropriateness of the technology has now taken a more public tone: http://gas2.org/2016/09/17/spat-mobileye-tesla-goes-public/

    "it said its chairman, Amnon Shashua, raised concerns with Tesla in May of 2015 — 6 months before the Autopilot system was activated. Shashua reportedly warned Elon Musk personally not to allow Autopilot to operate without a hand on the wheel. It further says Musk assured Shashua that Autopilot would in fact be “hands on.” But the system Tesla rolled out in the fall allowed drivers to remove their hands from the wheel for extended periods of time under certain circumstances while the cars took control of steering and other vital functions."

    Musk, of course, disputes this
    http://arstechnica.com/cars/2016/09/mobileye-and-tesla-spat-heats-up-as-both-companies-trade-jabs/
    and claims that MobilEye wanted to stop Tesla in-house development
    https://www.engadget.com/2016/09/16/tesla-mobileye-response/
    -- which is, of course, not 100% accurate given also the side connections between Tesla and NVIDIA-- the dashboard is powered by NVIDIA-- which have a well advanced platform in their Drive-PX...
    Anyone really believe that Tesla could develop something to rival the EyeQ4, Drive-PX2 or .. (I am a real fan of GPUs and NVIDIA but MobilEye has a pretty big lead here and to suggest that a shoestring group at Tesla would leapfrog I'd take as little more than smoke and mirrors to distract from their own ego failings) ?

  • Google self drive car involved in a collision because the driver of another car jumped a red light

    https://www.theguardian.com/technology/2016/sep/26/google-self-driving-car-in-broadside-collision-after-other-car-jumps-red-light-lexus-suv

  • Jumping red lights? Must've had a cyclist driving it.

  • Thinking of doing their self-driving car course-- anyone here also get accepted?

  • Person transport by rail and self-driving cars are hardly substitutions but rather complements. The wave of self-driving will, I think, hearald a new wave of rail as personal ownership of motorcars goes the way of the horse drawn buggy. Cars won't take one from London to Cambridge. Trains-- or self-driving buses-- will.

    Lacking the demand for parking places will mean that few if any motorcars will be private but find usage patterns like taxis. More problematic, I think, will be transport of goods as rail will find it also difficult to complete with fleets of linked road trains (mega-trucks). Railways are currently not doing well with freight and I suspect the conflicts will increase.

  • I doubt this very much. 'Self'-driving precisely levels out one key advantage of rail travel today, that people can work while they're on the train. While ownership of cars will probably move away from the individual, it's a fairly well-publicised fact that the car companies see the future of their business models in building cars, not selling those, and instead charging for trip hire. The aim is to undercut public transport and to effectively privatise it. This would obviously not be possible in the inner cities, but outside them it is a distinct possibility. Obviously, it's all speculation at this time, but I think this is a much more likely direction for it to go in.

  • Outside the cities and commuter routes to/from them most train services are crap though. One train an hour (which may or may not turn up depending on whether the train company has bothered to get enough drivers for the day) which proceeds to stop at every tiny station between you and your destination. People will need no encouragement to move away from them.

    I see express train routes between the cities and then self-driving cars picking up at the station and radiating outwards.

    This isn't going to change until all drivers are off the road and motorways move on to supporting convoys of cars at 150mph+ which I can't see happening in the near future.

  • I'm sure someone (probably in this thread, my memory is shite) said freight on trains is no good because the railways make too much off the sheeple using them whereas all of us tax-paying idiots are subsidising trucks to deliver stuff via roads, hence trains struggle to compete? Don't know if true.

  • doubt this very much. 'Self'-driving precisely levels out one key advantage of rail travel today, that people can work while they're on the train.

    As with taxis today, logistics will, I think, place strong cost demands that the autonomous motorcars don't venture too far off from their service hubs. Calling a automonous car to drive you from London to Birmingham will, I suggest, demand a significant surcharge to offset the demand and utilization differences between having cars ending up in Birmigham where they can't take passengers in London. There will also be issues, I think, with commuters as they tend to be in single directions-- morning, for example, into London and evening the other way-- thinking here of places like Luton, Crawley, Reading etc.

  • I'm sure someone (probably in this thread, my memory is shite) said freight on trains is no good because the railways make too much off the sheeple using them whereas all of us tax-paying idiots are subsidising trucks to deliver stuff via roads, hence trains struggle to compete?

    Correct. That is why privatization of rail transport was a very nasty idea. Unless one demands that commercial use of the roads is priced to reflect costs-- which would also have horrible side-effects on transport to areas not agglomerated, increasing the trend of urbanization-- it is nearly impossible for rail freight to compete. Given costs and latency of adapting to changing freight demands, rail will increasingly find it difficult to make a profit here. Megatrucks and autonomous land-trains will greatly expand these differentials. New technology is a freight logistic game changer. These trucks are liberated from EU driving time and rest period directives and can shift from convoys to highly efficient self-adjusting swarms.

  • As with taxis today, logistics will, I think, place strong cost demands that the autonomous motorcars don't venture too far off from their service hubs.

    Why is that an issue? To Uber it doesn't matter if self-driving car A is in Birmingham and car B in London or vice-versa, there's no differentiation between the vehicles. With taxis today the driver needs to go home at the end of his shift. There's no reason why a car has to (assuming it is owned by a company with a national presence).

  • Might have to stop for a few recharges but maybe that can be coordinated with the passenger if/when they need a piss/leg-stretch.

  • Why is that an issue?

    There can be a number of issues. As with rental cars I think it could lead to imbalances in utilization and costs to reset which need to be recouped. Should there be a particularly strong demand, for example, for people to go from A to B but not B to A and where demand is much higher in A than B. It can mean that too many vehicles are collecting in B where they have much lower utilization than in A, demanding perhaps that empty cars must be moved back to A. Taxis and rental cars have surcharges for this.

  • As with taxis today, logistics will, I think, place strong cost demands that the autonomous motorcars don't venture too far off from their service hubs. Calling a automonous car to drive you from London to Birmingham will, I suggest, demand a significant surcharge to offset the demand and utilization differences between having cars ending up in Birmigham where they can't take passengers in London.

    I don't think it'll be sold that way around, but rather that 'you can grab a bargain if you take a car back to London'.

    There will also be issues, I think, with commuters as they tend to be in single directions-- morning, for example, into London and evening the other way-- thinking here of places like Luton, Crawley, Reading etc.

    Sure, excessively tidal flows already exist, of course. As I said, in London, public transport will persist, as it simply moves far more people. However, self-driving cars will play a greater part because they eliminate one of the biggest capacity constraints in car travel, when each driver accelerates slowly, with a gap after the previous driver. Self-driving cars can, at least in theory, move off as one big block, meaning you get far more through a junction in one go. That would certainly increase 'private' (probably car company-owned) motor traffic.

  • Best bit is about all this is that if the robot cars get it wrong I can throw a brick through the windscreen without having to deal with that annoying squishy bit in the driver's seat.

  • https://www.youtube.com/watch?v=jrF1g4za_7I

    https://www.youtube.com/watch?v=xv-e5-uvemU

    How a self-balancing motorcycle removes the need for a helmet-- also how her protective pants can be reduced to fishnets-- is beyond me.. Cool, sexy video though...

  • Today I met some of the BMW electric motorcycle people at a tradeshow on future mobility. Being shown some electric kickboard, I asked "where is the Next 100"? Their answer? Nowhere.. just a dream..
    And like dreams, I guess, "So safe that you don't need to wear a helmet"...

    (by comparison


    was real)

  • You won't be allowed to use your self-driving Tesla for Uber or Lyft http://www.theverge.com/2016/10/20/13346396/tesla-self-driving-ride-sharing-uber-lyft

  • theverge.com/2016/10/20/13346­396/tesla-self-driving-ride-sharing-uber­-lyft

    Interesting watching the guy testing the Uber SDC (in this link). he seemed nervous, had the urge to step on the gas, and took control of the car a number of times. Sure, its really new tech but I suspect many drivers will take a while to gain proper trust in the things.

  • And the luddite opinion of Toby Young is completely unsurprising...
    http://www.spectator.co.uk/2016/10/how-driverless-cars-will-make-your-life-worse/

  • Post a reply
    • Bold
    • Italics
    • Link
    • Image
    • List
    • Quote
    • code
    • Preview
About

Robocars - Autonomous Drive, Self-driving, Driver-less cars

Posted by Avatar for hippy @hippy

Actions