Stories
Slash Boxes
Comments

SoylentNews is people

posted by on Wednesday April 19 2017, @04:21AM   Printer-friendly
from the someone-has-to-be-first dept.

Einride, a company based in Gothenburg (Göteborg), Sweden, has a vision that lowers these hurdles slowing the adoption of both alternative energy and self-driving technologies in hauling. The key change? Take the driver out of the vehicle with a hybrid of self-driving and remote control.

Einride plans to have their driverless (windowless, even) T-pods plying the route between Gothenburg and Helsingborg by 2020. The 7-meter (23- feet) long vehicle can carry 15 standard pallets and up to 20 tons. The trucks roll through their highway distances in fully automated mode. But when they near population centers, the T-pods can be put under remote control, with a human managing the navigation.

With no paid personnel on board to be bored and useless during long charging cycles, electric motors begin to make more sense. The T-pods can travel 200 km (124 miles) on a single charge, and stops at charging stations add little to the overall costs of haulage compared to traditional rigs that have down-time during driver resting periods. Remote drivers can simply switch their attention to a different vehicle when one T-pod stops for recharging. Which is a good thing, because even the run up and down the Swedish coastline between Gothenburg and Helsingborg may be a bit out of range without a top-up along the way.

Maybe all those hours playing Starcraft did not go to waste after all--perfect training to be an Einride operator.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 1) by tftp on Wednesday April 19 2017, @05:02AM (6 children)

    by tftp (806) on Wednesday April 19 2017, @05:02AM (#496148) Homepage

    Remote control will have a significant delay. But for safety it requires transferring real time data (4K video) with latency under, say, ten milliseconds, camera to screen. Not sure how this can be achieved, given that many monitors need most of this time just to refresh the screen.

    Who would want to drive a truck in "population centers" if the visual lags the reality by half a second, sometimes stops (wireless!) and the commands reach the truck another half a second later (or not reach it at all)? It would be safer to just go 100% automatic.

    Looks like yet another startup collecting investors' money to build something that not only violates the laws of nature (there is not enough bandwidth for this under 60 GHz,) but also is a stopgap solution that will be obsolete in 5-10 years, probably before the investment starts generating profit.

  • (Score: 0) by Anonymous Coward on Wednesday April 19 2017, @05:27AM

    by Anonymous Coward on Wednesday April 19 2017, @05:27AM (#496156)

    I was thinking about this, and I think the delay can be handled by self-driving tech.
    The problem, at least as I see it, is that self-driving tech is not good enough to navigate complicated roadways. But it is good enough for low-latency things like "brake for the child that jumps into oncoming traffic." All high-end cars and most mid-range cars already have much of that capability as standard (lane keeping, automatic braking, etc). So let the computer handle the fast-response stuff, while the remote driver makes the higher-level decisions about which turns to take, etc. It would probably require a different set of controls than the traditional steering-wheel and accelerator/brake pedals. Maybe a joystick with an on-screen representation of the path the vehicle is on course to follow.

  • (Score: 2) by Scruffy Beard 2 on Wednesday April 19 2017, @05:32AM (2 children)

    by Scruffy Beard 2 (6030) on Wednesday April 19 2017, @05:32AM (#496157)

    The "Fragile" comment below mostly got it right, but I think mixing human and automatic input may work.

    The vehicle just needs proximity sensors and logic to automatically stop (fixing the latency problem). 4K resolution may not even be needed in built-up areas due to the limited speed (fixing the bandwidth problem). The operator is essentially responsible for object avoidance and path-finding.

    Still don't think it is a good idea, but not sure impossible is the word for it.

    • (Score: 2, Insightful) by tftp on Wednesday April 19 2017, @05:55AM (1 child)

      by tftp (806) on Wednesday April 19 2017, @05:55AM (#496159) Homepage

      I think mixing human and automatic input may work.

      Yes, but the easiest way to achieve that is by having human drivers who jump behind the wheel at a certain transfer station and complete the route into the city. As long as the truck comes charged for the remaining miles, there is nothing to gain from having remote drivers, and plenty to lose. Why are they removing manual controls? Are they going to save money on the steering wheel, windshield and a chair? Everything is already drive-by-wire, plug your controls in and go. Just the certification of the remote control system will take years: as others said, it *is* a safety critical system, and it will be necessary to prove (not just show once) that the combination of the automatic and remote controls will not incur additional risks. In particular, "just stop" may be a terrible answer to loss of communication if other drivers do not expect the truck to stop on a ramp or on a freeway or in many other places.

      • (Score: 0) by Anonymous Coward on Wednesday April 19 2017, @04:24PM

        by Anonymous Coward on Wednesday April 19 2017, @04:24PM (#496392)

        Another approach, if they want to keep the T-pods cabless, is to have a follow-the-leader mode where you slave each one to the one in front, and slave the front one to a human-driven vehicle.

        This can be particular sensible if several T-pods are going to the same place, but also works if you can set up a chain of 3 or 4 T-pods, and drop them off at various points on a more-or-less direct route. It doesn't make any kind of sense if you're leading one T-pod at a time

  • (Score: 2) by bradley13 on Wednesday April 19 2017, @06:35AM

    by bradley13 (3053) on Wednesday April 19 2017, @06:35AM (#496166) Homepage Journal

    Don't exaggerate the technical requirements. 4k resolution? 10ms latency? Nuts. The average driver reaction time is on the order of 500ms, for which 100ms latency would be fine - ugly, but fine. To fix the ugly, make it 50ms. A 4k resolution is also unnecessary; A standard monitor resolution would be more then adequate.

    Bandwidth is not necessarily a problem - if this becomes widespread, it will get dedicated spectrum. Moreover, you don't need to control all vehicles all of the time. You only need human control when the vehicles are at the end of a defined journey (off the major roads, say), or when a vehicle yells for help because something unexpected happened. In the latter case, the vehicle can stop and wait for human help.

    What one may want to rethink is the way information is displayed. We are used to looking out of our windshield, but an eagle-eye view from the top, showing the situation around the entire vehicle, might be more useful. Or, perhaps, some combination of the two.

    --
    Everyone is somebody else's weirdo.
  • (Score: 2) by davester666 on Wednesday April 19 2017, @08:24AM

    by davester666 (155) on Wednesday April 19 2017, @08:24AM (#496194)

    Not only that, but it will become a race to the bottom (and past), in reducing the number of drivers needed, and then making individual drivers work longer and longer hours by offsetting trucks by less and less time. You have to truck X from the hwy to the dock in 10 minutes, because there's another truck to get from the high to the dock arriving in 10 minutes. And you're fired if you fail, we'll find someone who can do it.