[Nokinside]

Remote operators solve these issue.

Once you have autonomous cars that driving safely, but can't manage complex situations like you describe, you delegate those for remote pilots that are allowed to operate car in slow speeds. You need 5G network coverage with mission-critical features (mcMTC) to archieve that. BLER 10^-6 and E2E latency < 5- 10 ms. Construction work crews might be required to erect 5G mini cell tower before they can start working to make sure that traffic goes smoothly.

Taxi fleet of 10,000 vehicles might need only 100-200 remote operators to manage the fleet. That kind of reduction in workforce provides huge savings.

[leoc]

> You need 5G network coverage with mission-critical features (mcMTC) to archieve that. BLER 10^-6 and E2E latency < 5- 10 ms.

I do wonder if that's a factor behind Musk's push into low-orbit satellite Internet.

> Taxi fleet of 10,000 vehicles might need only 100-200 remote operators to manage the fleet. That kind of reduction in workforce provides huge savings.

Even if all mileage was human-driven there would be very large benefits if you could really consolidate taxi drivers in call-centres for remote driving. No need to transport or preposition drivers and much less trouble estimating demand.

[Nokinside]

Starlink can provide 25 to 35 ms latency from low orbits, so I don't think so.

[leoc]

Even 70ms isn't all that long compared to the the official (let alone real-world) thinking-time estimates for braking http://www.brake.org.uk/component/tags/tag/thinking-time , so it might not be an unacceptable delay, at least if combined with professional drivers and reduced speed limits and/or failsafe locally-controlled AI braking.

[masonic]

  fleet of 10,000 vehicles might need only 100-200 remote operators

It won't scale in a predictable way. Let's say there's a major event in NYC (natural disaster or unnatural). You may suddenly need 700 operators at the same time just to deal with NYC and environs.

[vasilipupkin]

Let’s say there is an event in NYC. Within milliseconds of the event, all the robot cars can be notified and their human drivers can take over, only the cars immediately in the vicinity of the event need robot operators. It’s not straightforward, but it can be done.

[vonmoltke]

Thus requiring the car to always have a licensed driver in the driver's seat. Basically, exactly what we have now with the safety drivers.

[ghaff]

Even if you assume requiring human takeover to be a relatively uncommon event (whatever that means), as soon as you posit it as something that will be needed from time to time, you've significantly constrained the car's usage models. You now must have a licensed, unimpaired driver in the car at all times. Even if they don't have to be paying attention, this means no empty cars, no unaccompanied children, no "driving" home from the night out, etc.

[vasilipupkin]

Sure, but maybe we can start there. I would certainly buy a car that could drive itself a significant percentage of the time

[ghaff]

Oh, I would too assuming it were relatively affordable. I'd be pretty happy with one that even just let me doze off when highway driving in a limited set of weather conditions.

I was just pointing out that, if you can't guarantee you won't need to handoff to a physically present driver, then there are a lot of things you can't do with the car even if needed interventions are just an occasional thing.

[mlonkibjuyhv]

Yeah... Used to work for a small start-up that had a product to basically automate a switchboard used for elderly care. It was fun when the need for manual operation suddenly came around. Didn't have the manpower, nor the actual switchboard.

[Nokinside]

You reduce the number of vehicles available in those rare cases.