Hacker News Reader

Comment by willis936 | Hacker News Reader

What about an autonomous semitruck that could run 24 hours a day except for unloading/reloading cargo, refueling, and maintenance?

View on HN

Trains. The fact that long haul semi trucks are a thing is a failure of infrastructure.

Edit: as some have pointed out, it’s more a failure of policy.

Trains are large and inflexible. Short of having a city-block resolution freight train network at every population center, you'll need something flexible for the last mile of delivery.

You’re misunderstanding.

The only reason why you’d need on-road recharging for hypothetical autonomous semi trucks is if they’re regularly traveling outside the range of their batteries, which is probably in the 400-500mi range.

In a well designed infrastructure system, that shouldn’t happen often. Instead the hops of hundreds of miles should be handled by trains, leaving semi trucks for the last dozens of miles from train hub to area of need.

Some long haul trucking will happen, but it should be a slim minority of cases.

Adding an extra step (rail hub to short haul truck transition, potentially at both ends of the trip) adds cost. This cost is (and will be for the foreseeable future) less than having the less efficient semi-truck do the entire trip at least for the kinds of loads that are currently carried by semi-truck. JIT inventory systems also don't play nice with rail.

Semi trucks are cheaper because they receive a $125B+ yearly subsidy in the form of road maintenance. Semi trucks do 99% of the damage to our public roads, but only pay 35% of the cost.

I suspect that rail would be much more price competitive if we took that subsidy away.

You’re right about JIT, but we’re also seeing the downsides of JIT infrastructure during this pandemic. I suspect that we’ll see a push for more redundancy in our supply chain for reasons completely unrelated to trucks vs. trains.

Assuming we're resurfacing our roads once every 5 years, if we banned trucks it would take 100x longer to accumulate the same amount of damage, so we'd be able to resurface only once every 500 years?

I would expect there to be a significant amount of damage done just by the elements, independent of usage.

There’s also the fact that roads are built up to trucking standards. Without trucks the roads would be thinner, reducing cost and pollution (concrete is a major source of CO2).

Roads are mostly made from asphalt and mother nature would still destroy them in well under 100yr in most places.

Highways are much more relevant when we’re talking about trucking, and highways are largely concrete.

This is utterly false and in a trivially verifiable way in this age of street view.

Raised highways sometimes have concrete decks because the strength can be attractive but the overwhelming majority of highways are asphalt. At or below grade highways are almost always asphalt. Tunnels are often concrete but those are a vanishingly small minority of miles.

If you’re going to call something “utterly and trivially verifiable (sic) false” you should actually check. Interstate highways use more cement than asphalt.

“The analysis shows that a total of approximately 1.5 billion metric tons (Gt) of aggregates, 35 million metric tons (Mt) of asphalt, 48 Mt of cement, and 6 Mt of steel is in place in interstate highways.”

https://pubs.usgs.gov/fs/2006/3127/2006-3127.pdf

AFAIU, the original Interstate Highway System was all concrete. These days states usually resurface heavily trafficked segments with asphalt (of various varieties) atop the old concrete. Apparently, at least as of 2006 60% of the system was still concrete pavement: https://pubs.usgs.gov/fs/2006/3127/2006-3127.pdf Which is not surprising as IME long segments of the IHS are often still concrete pavement, especially in the South and West; it's only in urban areas where asphalt surfacing seems ubiquitous.

I'm not sure how new segments are built today, but they may very well still use concrete underlayment, or at least something more substantial than loose aggregate. Road tech these days is way more complicated than just concrete v. asphalt. I would guess that for remote segments, and given the load requirements, it may still be cost effective to keep using concrete surfacing.

Sure. I don't necessarily disagree with your conclusion that trucks undercontribute to road maintenance costs. I just don't think those numbers are sufficient to prove it.

You might still want want "in flight" charging even for shorter trips just to keep trucks moving but I guess the world where load/unload times aren't enough is pretty unrealistic.

The inflexibility of trains can be a feature in some cases. People like to live near subway and commuter rail stations partly because they know the route and service level are much less likely to abruptly change on them than a bus route would.

They’re also more immune to congestion. Back when I lived in Chicago and used public transit, one would strongly prefer the rail over buses when possible, except in cases where the bus was an express route.

100 trucks running right up next to each other is a train. I could imagine train operators converting their right of ways to a system where each train car can be autonomous and can leave the train and track without disruption. As the train comes by an exit location (station), the cars that need to get off just come out of the train and leave while the train tightens up again for aerodynamic efficiency and continues on without even slowing down.

These single tracks could even allow individuals to run their own single/double cars on them if the track was not to crowded and premium price was paid. With automatic vehicles, a reality on such controlled tracks today, our current infrastructure for transportation is definitely a failure compared to what it could be.

I don't see that happening. Train cars are incredibly simple things that have barely changed in many decades, probably because it's hard to introduce changes when all the cars need to be compatible with one another. Train cars don't even have electrical hook-ups, just air brake lines.

True for freight, but not true across for all rail. High speed rail in particular is usually a different gauge and has electric hookups for all cars, because every wheel needs to be driven to hit those speeds.

Trains are far more efficient, less subsidized, safer, less polluting...

https://grist.org/article/freight-trains-19th-century-techno...

Even if you electrify trucks the tire and brake wear and road degradation contribute an enormous amount of PM 2.5.

Sure you'll need trucks for the last mile but every ton of freight we can move to rails or waterways is a victory for humanity.

I used to think that too, but

- economically trains are a monopoly. This point dwarfs all others.

- we don't have two-directional rails everywhere

- electric trucks can change the economics and mitigate the pollution of trucking

Electric trucks will still emit a rather large amount of pm2.5, and are still heavily dependent on both direct and indirect federal subsidies.

The thing that really feels possible to me is long haul autonomous trucks only on the highway. They can transport themselves between depots and then an actual driver does the tricky in-city transit and delivery/pickup.

I would be extremely opposed to this on efficiency grounds.

Semi trucks are awful. They’re horrible polluters, and they represent 99% of the road damage being done, despite only kicking in 35% of road upkeep cost. This is a drain on the public purse in exchange for more pollution and more accidents.

A far more sane policy would be to shift as much freight over to rail as possible, and reserve semi trucks for the last miles between rail head and customer.

Also, rail exists today. Self driving trucks don’t.

A diesel electric duorail truck would be great for large freight loads.

Maybe you’re onto something here; what if we just rebrand trains to appease the car centric average American?

Reply on news.ycombinator.com