A new service in Los Vegas called Halo allows customers to summon a rental car to them. The car is remotely driven by a human over the 5G data network, and dropped off for the customer to use as a rental car driven by the customer. When the customer is done, the remote operators take it back to base or the next customer.
This realizes a form of something I outlined in 2008, the “Whistlecar.” I envisioned this as an early form of the robocar, for a world where it was too difficult to make a car that could drive passengers in full safety on fast and complex roads, but possible to make a car that delivers itself, sticking to simpler roads and situations, and putting no passenger at risk. In that situation, it is possible to make a fairly useful service. You can either view it as a carshare/rental car that delivers itself to you and lets you drop it anywhere, or as a “Uber” style service where you do the driving.
This isn’t quite that, because the driving is done by a remote operator. It was an error to presume that simple unmanned robocar operations would occur before data networks that were good enough to use for remote control driving. In fact, it’s possible that the robocar leaders could make the whistlecar (so named because it was like Roy Rogers’ horse, which he could whistle for, and it would come to him to ride) but don’t wish to because they have much grander ambitions with the full robotaxi.
Nonetheless, absent the robotaxi, it’s quite useful. Existing carshare usually requires you to walk some distance to get your car. The supply and location of cars may be random in the case of stationless carshare like “Share Now” (which at least you can “return” anywhere in a large zone if you can find parking) or it may be fixed, as in ZipCar, where you must return the car to where you got it. Neither of those is very convenient, but Halo solves this as long as they can get you a car reasonably quickly.
Alternately, this can also be viewed as a great way to use remote drivers. Having full time remote drivers — ie. ones who drive even while you are in the car — is only marginally more efficient than a regular taxi. You still need to pay a human driver for time to get the car to you and to drive you around. However, there are these advantages:
- No human is needed while the vehicle is sitting idle, waiting for a fare. Likewise, the cost of having the car wait for you after it arrives is minimal.
- Drivers can work in one place, anywhere within hundreds of miles, as long as they have quality networking. They don’t have to drive far from home and need to commute back home on their own dime at the end, as many Uber drivers do.
- There’s more privacy in the car and room to carry one more person, though there are downsides if the customer damages the car with nobody to supervise.
With a service like Halo, it can be much cheaper because the customer does the driving. While cheaper, that of course does not let the customer do other things on the ride. Nonetheless, for longer rides, where the Uber price of $2/mile can get very expensive, this can be a big win.
A more difficult question is whether the cellular networks are really up to this. It’s an open question if remote driving, even with lots of bandwidth and low latency, is sufficiently safe. Presumably the cars contain extra sensors and computing to try to fail-safe if the network connection is interrupted, even for a short time, but that’s not a well tested system. On the other hand, the remote drivers get a wide “wrap around” view which can, in theory, given them a better chance to see everything than a human in the car, even with mirrors. Time will tell.
T-Mobile is an official partner here, in order to show off their 5G deployment in Las Vegas. 5G is often spoken about as a magic bullet, but it’s really just the latest LTE combined with microwave short-range microcells for extra bandwidth and some latency efforts. It’s the latter that would be making the difference here. The cars themselves don’t need as much modification as a robocar, including a LIDAR, though it could help them to have that. The remote drivers will know not to drive the cars in any areas of poor network quality, even if that means driving the long way around, to keep outages to a minimum.
Many other potential advantages were included in the Whistlecar vision. One of the most important was the ability to deliver the “right car for the trip” the way a robotaxi can. Ideally, the Halo fleet would consist mostly of narrow, single person cars rather than sedans and SUVs. That’s because 80% of trips are one person across town. If a service used such cars it would not only be much cheaper (these cars cost a lot less to make) but also cause less pollution and congestion.
Halo apparently keeps cars in a depot. ShareNow cars are parked anywhere in a city, and Zipcars are in specific special parking spots. If the trip from the depot is long, that could degrade the quality of service. Ideally a service like Halo would want to have lots all over the city, like Zipcar, or do a contract with the city to use available spots like ShareNow. It will be interesting to see where this goes.
Remote Operation
Several firms are pursing remote operation. Most robotaxi firms, such as Waymo, have remote operators centers but the remote operators do not drive the vehicles by remote control. Rather they just given them strategic advice, like where to turn or what lane to take, leaving it up to the robot on the ground to work out the actual steering and accelerator. Aside from the issues with remote driving and latency, it’s still the case that there isn’t reliable enough network bandwidth to depend on this over a large service area, but remote advice can be done over much poorer bandwidth.
At the same time, remote advice often involves a vehicle stopping to wait for the advice, which can mean bad road citizenship for cars. (Sidewalk delivery robots can stop much more quickly and can be easily walked around, so they can rely more on remote advice.)
Waymo famously had a problem with their remote advice system in Chandler, Arizona, when a remote operator gave the wrong information to a vehicle, causing a comedy of errors.
Full time remote control has some benefits over a human driven taxi, but it still requires a paid human in the loop. Remote advice can be given half the world away, but remote control needs to be done by somebody closer and probably similar in wages to a local driver. Even so, part time remote control driving can be useful for certain circumstances, including carshare as above:
- Getting a car out of a jam faster than remote advice if the bandwidth is available.
- Taking passengers outside the robotaxi service area, both to remote areas and internal streets not handled by the software, thus greatly expanding the usable service area.
- Handling known “tough zones” on a robotaxi route, temporarily taking over from the software when planning a route through a zone that is known to have issues (construction, current conditions, current traffic, etc.) if the zone has good connectivity. (Otherwise route around it.)
- For trucks, taking them from off-ramps to depots on city streets, so that the self-driving system only needs to handle highways.
- Driving electric cars at night from parking spaces to charging stations (until they can do this on their own) so that people need not install charging nor worry about it.
- On-demand driving service for otherwise human driven cars with the necessary hardware. We earlier debated if VW’s proposal to sell self-driving by the hour at $8.50/hour would work. What about remote driving at $20/hour? Certainly some would pay for it.
Starlink
An interesting possibility could come from services like SpaceX Starlink, which promise low latency connectivity at high bandwidth over the whole globe, excepting tunnels and possibly some urban canyons. Starlink’s receiver is currently a bit expensive and power-hungry, but could be justified to solve the above problems for robotaxis. It will probably get cheaper. This would make the most sense outside dense cities, both because reception will be good and usage density will be low, making the bandwidth spare.
Of course, Starlink is run by the same man as runs Tesla. That could make it more readily available to any Tesla based system, and also make it not available to competitors. However, other competitors to Starlink are in the works. Starlink will also offer fairly low latency even long distances away, allowing the use of remote drivers in lower wage areas. (Starlink will eventually relay signals through space at the true speed of light, rather than the slower speed of light in wire/fiber.)
"do it" - Google News
July 11, 2021 at 10:00PM
https://ift.tt/3yJqhr8
Halo Uses Remote Operations To Offer Summoned Carshare; Might Tesla Do It With Starlink? - Forbes
"do it" - Google News
https://ift.tt/2zLpFrJ
https://ift.tt/3feNbO7
Bagikan Berita Ini
0 Response to "Halo Uses Remote Operations To Offer Summoned Carshare; Might Tesla Do It With Starlink? - Forbes"
Post a Comment