Hacker News Reader

Comment by jvanderbot | Hacker News Reader

> drivers leave the industry

is basically the same thing as

> there's a shortage

If anyone wants to work on this problem with me without solving the self-driving truck / Aurora problem outright - I think the answer is something like remote operation, level 3-4 autonomy for highways, and nice comfy air conditioned driver hubs in cities so nobody has to be away from home for days at a time. Boom - all the drivers come back and get paid decent wages to drive trucks. Ideally the per-truck subscription fee for the service is an add-on to the cost of doing driver business.

View on HN

Remote operation on public highways is a dead end. There is no wireless data network with the latency and reliability guarantees necessary for safety. Particularly when it comes to tunnels, canyons, and heavy precipitation. And let's not have any ridiculous claims about how in case of a network connection failure the truck can just autonomously pull over and stop; that's not going to work safely for descending I-70 in a snowstorm.

There may be some limited use cases for remote operation on specific local routes where sufficient network access points can be provisioned.

It’s not a dead end, it’s a solvable problem but the issue is that no one wants to pay for infrastructure.

You can't do it with autonomy, and you can't do it with remote piloting, but you can do it with a mix of both. That's my assertion. Spending 20 hours looking at Nebraska / Illinois farm fields is not the use case for remote operation - and as you say, you lack the infra for that anyway. And navigating traffic or interacting with distribution center logistics is not the case for autonomous operation - it's an infra nightmare to get a autonomous vehicle to integrate with all that radio-voice-comms madness. Even OTR drivers just want to get that part of the journey over with and get back to the hotel.

Having a pod with local drivers near major hubs, for example, means the drivers take over when the trucks get close enough for it to matter. It wont' work for delivery to, say grocery stores (which can be local driving anyway), but it will work for center-to-center transport ala between Walmart hubs. It's the long haul OTR trucking that has high attrition / people shortages, because you're away from family for so long. That little bit can be somewhat autonomous, with handoff of remote operators only near hubs. Think air-traffic-control for trucks and hubs, except probably more hands on than just telling it what to do.

Like harbor pilots for ships, where someone comes out to join the ship to help navigate through potentially dangerous waters.

Even without full autonomy, truck platooning has been shown to be effective in places where they have been tested. One human driver and "autonomous" followers. These truck platoons could be the solution to long haul deliveries while individual drivers still handle the last mile delivery and city navigation.

https://projects.research-and-innovation.ec.europa.eu/en/hor...

"> drivers leave the industry

is basically the same thing as

> there's a shortage"

No, it's not. It means the drivers are not being paid enough to stay, as they explicitly say is the reason they're leaving.

>If anyone wants to work on this problem with me

Sure, here's the solution: pay the drivers more. Problem solved.

> drivers are not being paid enough to stay

therefore

> drivers leave the industry

therefore

> there's a shortage ... at the going wage

I don't see how we can argue over the problem statement. We can argue over the solutions, of which there are infinite, but at least two:

1. increase wages

2. decrease shittiness of the job via ideas like mine

Or why not both? The only dichotomy here is a false one you've introduced.

Because there's no problem. There's not "shortage", it's not that there just aren't enough people to be drivers. There are plenty, employers just don't want to pay enough to attract them.

>We can argue over the solutions

No, we can't, because there's no solution needed to this non-problem artificially created by employers unwilling to pay enough to attract and retain employees.

Your ideas aren't needed except by the employers trying to perpetuate their false self-made "shortage", all to avoid just paying employees more.

The only actual problem is uncritically accepting the narrative of the employers as you are doing here.

> There's not "shortage", it's not that there just aren't enough people to be drivers. There are plenty, employers just don't want to pay enough to attract them.

OK, well, let's not mince definitions. For the purposes of my thought process, a shortage is defined as "Number of people we want to do a job" minus "Number of people who do that job". The semantic difference between that and what you said is so small as to be negligible to me and doesn't preclude "increase wages" as a solution. I think that's fine for you to have a preferred solution or even a different preferred definition.

No, you're just wrong. You're carrying water for the employers who want to squeeze their employees.

Again, there is no actual problem here, just an artificial one created by employers unwilling to pay their workers and crying about it. Don't buy into it.

It's not about employers in this case. If we think being a truck driver is an important job we can subsidize truck drivers federally, via taxes.

At the end of the day, society (in the form of government regulation) gets to decide what is and isn't an important job (i.e. government subsidies), and we the people can certainly put our thumb on the scale.

People collectively decided truck drivers were overpaid [relatively] in the 1970s, hence the deregulation of the trucking industry which led to a collapse in trucker's salaries. No use blaming employers here - it was decided by the government (representing the people of the U.S.)

> It's not about employers in this case.

In this case it is. The discussion is about how employers have a dream that isn't being realized. They are metaphorical 10 year old-boys drooling over the idea of owning a Ferrari, without the capability to actually buy one.

The only "problem" is that their dream is remaining a dream. Which isn't actually a problem. Dreams aren't supposed to be realized. They are meant to be just dreams. Who gives a shit if a 10 year old can't own a Ferrari? And that is what the parent is pointing out – that the '10 year-old cries' of business are meaningless. Let them dream, but it ends there.

If people of the U.S. want to give truckers more money, they are free to do so, but that is well beyond the topic at hand.

I completely understand and agree with your point. Just wanted to be clear. Not disagreeing at all.

Talk about rhetorical self harm. The person you're screaming at isn't far from you, but you refuse to acknowledge a premise.

Notably, you ignore their autonomous driving pitch is supposed to increase working quality of like, like many instances of automation (in theory).

What if someone can think drivers deserve better pay and ethical treatment and that self driving is a long term solution to a grueling job?

>but you refuse to acknowledge a premise.

Why would I acknowledge a flawed premise?

>like many instances of automation (in theory).

That (in theory) is doing so much work here it ought to be paid.

>What if someone can think drivers deserve better pay and ethical treatment and that self driving is a long term solution to a grueling job?

What if mythical beasts like unicorns existed? That would be pretty cool, I guess. Sadly, we'll never know.

How realistic are the trucking sims? Do you actually get gas and unload? Highway stuff and regular driving sounds good but backing a trailer up to Walmart video game style sounds frightening.

We can do that scary bit autonomously already: https://www.outrider.ai/

I'm not referring to "sim", I'm referring to remote piloting. Real time sensor and controls feedback from a truck that is only autonomous when it makes sense to be. We can pilot drones over 100s of yards in real time without a hitch. We just need a bit more range (few kms) via, say, the internet, and you can do this fairly easily. I don't see a research problem, I see a funding problem and an aversion to autonomous trucking b/c people think it's either completely infeasible, or completely solved by Aurora. I'm not sure of either.

And what does the driverless truck do when it loses its internet connection at 70mph on a freeway?

The use case does not include operating remotely at 70mph on freeways, esp because internet connectivity is hard over most the USA. That bit is mostly lvl 3/4, and is something that is addressed well by other companies. It's the long tail between lvl3/4 and all the scenarios that remain in which a remote driver solves the problem. Esp in logistics yards, which are chaotic, fast paced, and driven by voice communication.

OTOH we dock manned spaceships to the ISS video game style

That is a much easier problem than backing a truck up to a loading bay. Even landing an airplane is a lot easier if you have an ILS. Space and the air are mostly empty. A warehouse loading area is cluttered with other trucks, debris, and cargo. Although I could totally see this being done remotely with VR and a ton of cameras.

Agree with you.

Just to add spice the convo: Backing into a loading bay is quite easy, autonomously. Once you stage the truck you can push a button and just watch it go. See: https://outrider.ai (which is where I work, and we do all the staging _also_, but I'm actually interested in "everything else" as well). There's a lot of very human-ish problems to get to that state, that I think remote operators can really help with.

Pretty good for the expensive ones. They are not video games though, and so they are not making the shortcuts games do to make them fun. I've driven a few (snow plow, excavator) and they are very hard to control - then I see someone who drives that machine professionally get on: they get a near perfect score and comment about how real it feels. Note that part of the sim is the controls: they take the cab from a real machine and replace the windshield with a screen (for version 1, often version 2 starts to take away parts, but everything you touch including the seat is the real thing).

I've long thought that to renew your drivers license you should have to spend a day every few months in a simulator to prove you can drive safely. Do you maintain a safe following distance? Can you maintain the speed limit? Do you stop for pedestrians? Do you zipper merge correctly? Here is a situation where someone else screwed up and you will crash, can you choose the best crash? Here is an icy road, can you control the car? there is a lot that modern simulators can do but few people have used them.

I agree with you on the DL renewal based on a sim.

Interestingly: The base CDL is a one and done skill test. There are no CPEs like with other regulated industries (nurses/teachers/investment advisors) - but there is a health and fitness check.

The difficult part is actually the last mile/few miles in trucking. Open highway driving is a significantly less complicated problem than navigating down side streets then backing into a warehouse to complete a delivery.

Freight tech companies have been talking about automating every BUT the last mile/last few miles for years. Have a stable of drivers who you have on hand to take over the last hour of a delivery. Automate everything else.

The nice thing about this is also that truckers would be able to go home everynight which would make the job more attractive

>If anyone wants to work on this problem with me without solving the self-driving truck / Aurora problem outright

In the US, we un-solved this problem decades ago when we abandoned railroads for subsidized auto development.

Reply on news.ycombinator.com