Hacker News Reader

Comment by jusben1369 | Hacker News Reader

So you could build dedicated infrastructure that within 25 years might meaningfully move people and/or goods between two points at 500 mph to 800 mph via a single tube. Or you can move to complete self driving cars and trucks that without the risk of human error could travel on the existing infrastructure (in many case 2 - 5 lanes wide) at 200 - 300 mph within the same timeframe or shorter.

I think self driving electric vehicles on the nation's roadways make the hyperloop at best something like Amtrak today in the US. Valuable and profitable only a few short routes/use cases.

EDIT: I do think we'll solve for 200 - 300 mph in 20 - 25 years. But insert 150 - 200 mph. The main premise is we're not going to build multi billion new infrastructure when a "good enough" solution is organically evolving. Same reason why we don't have high speed trains in the US.

View on HN

It's functionally impossible to drive at 250 mph for any meaningful period of time. Drag is a gigantic problem (as james may says "it's like driving through custard"), not to mention the fuel requirements to go that speed. One of the bigger problems is actually tires. The veyron's tires can only last ~20 minutes at those speeds and a new set is 20 grand. If you want to go those speeds you can't do it in the open air (unless you're actually flying) and you certainly can't do it on rubber tires.

Just allow cars to drive inside the hyperloop tube.

That's not solving the issue, which is using road instead of a rail-like system.

It only works in flying because you are getting up higher where the air is less dense.

Self driving cars, which aren't likely in that timeframe to do 200-300mph on roads that will exist, make rail and similar things, including hyperloop, more, not less, viable. They provide a solution to the need for a local transport component that doesn't rely on everyone owning a personal vehicle but still provides good local point-to-point transport to help enable the centralized long-distance arteries whose termini can't be convenient to every source and destination point without a good local network.

Self-driving cars that eliminate the need for personal car ownership (i.e. robo-Uber) are also unlikely to exist in that timeframe. What likely will exist is relatively slow travel on Interstate highways that don't require the driver to pay attention. Which will likely be interesting to a lot of people.

But the real benchmark for high speed rail (including hyperloop-like things) is how it competes with air on both total cost and time (the latter of which could be improved as a matter of airport policy and infrastructure). New York to Chicago in 8 hours--about half the current time--even at prices competitive with air--which Amtrak really isn't today--isn't going to be popular, especially among the business travelers who buy most of the tickets.

> on the existing infrastructure [...] at 200 - 300 mph within the same timeframe or shorter.

This isn't even remotely plausible. 250 mph in a conventional car is more of a party trick than a transportation strategy.

You can get a few very specialized cars up to these speeds on a few very specialized tracks, that's it. Even if you built a much longer straight section (you're not cornering significantly at these speeds) the question isn't going to be how many of them can you get from A to B, it's going to be what runs out first, fuel or tires. The range would make todays electrics look positively luxurious. Fuel efficiency would be low single figures.

You can't beat the physics here - you are fighting a velocity cubed term and it will always win (on existing infrastructure).

The few cars capable of driving over 200mph (none exist that can break 300, and they probably never will) are incredibly fuel inefficient. The Bugatti Veyron will empty its entire 26 gallon tank in 12 minutes at top speed, that's not sustainable for mass-transit.

I have a hard time believing that the infrastructure here in the Northeastern US could sustain vehicles regularly traveling at speeds in excess of 200 mph. Which means the cost would also require a great deal of overhaul to the existing infrastructure. It seems that many public works departments are having trouble keeping up with the demands of repairing roads for 40 - 60 mph.

If you could do a few hyperloops* that would connect Boston - NYC - Philly - DC, suddenly the strain on the highway system plummets. That gives self-driving cars even more room to speedily take people between locations outside of the reach of public transportation.

* I don't mean it would have to be hyperloops. Any HSR system would be beneficial.

Right but the challenge works the other ways. It wouldn't take much to support 150 to 180 mph. So Boston to NY which is more or less 200 miles becomes a 90 minute trip on a self driving car. So a good hyperloop gets it to say 30 - 40 minutes. Add to that all the approvals of laying one down in that heavily congested land ownership corridor vs the roads that are already in place. I just don't see it.

> It wouldn't take much to support 150 to 180 mph

Do you have a source on this? I did a little napkin math based on these suggested curve radii[1]; namely, I fit a function to the data points in this SageMath worksheet[2] and extrapolated. The fit a) looks good and b) seems to imply that the vastly increased minimum radius of curvature of turns would require significantly re-engineering the interstate highway as it exists today.

[1]: http://www.state.nj.us/transportation/eng/documents/RDM/sec4...

[2]: https://cloud.sagemath.com/projects/56f621e5-fd90-4a21-ba52-...

300 mph at sea level atmosphere on land is pretty tough to do efficiently, you're looking at ~18x the drag force of normal highway travel with only a ~4x speedup. Self-driving "trains" of cars can alleviate the drag force, but I haven't seen any 300mph "train"ed proposals that would work on existing road and bridge infrastructure.

Now imagine the 2 techs combined, hyperloops to link major hubs at the speed of sound and self driving cars for everything else.

> at 200 - 300 mph

What supercar are you looking at which is capable of this?

Yeah I thought the same thing. Don't the Veyron's tires wear out in like 50km at top speed?

Fast cars like that already exist, the main problem with those is that thanks to drag (air + surface), they become really uneconomical at higher speeds. 800 of the Bugatti Veyron's 1000 BHP only kicks in above 200 kph, maybe later - that's how much extra power is needed to squeeze out the last bit of performance to reach that max speed. Do that times the amount of cars, and you see the problem there.

Although it probably would work if the cars drove in a partially evacuated tunnel. But then you wouldn't have the downforce needed to get the cars through corners at those speeds. If you put the cars on monorails though, maybe...

You could call it the hypercarloop.

I would think fully electric buses on dedicated roadways would be cheaper and more flexible in the end. Automated driving is coming closer and the best uses are for those which travel well marked and maintained routes, hence long distance trucking, buses, and the like. It certainly would be cheaper than building any type of elevated rail system.

However with most transportation projects rail has way to much allure even when its financially stupid. Just look at the costs of some light rail projects in our cities and it never makes sense.

Maybe it's because I don't live in the US but I couldn't disagree more with this. Public transport is infinitely more valuable than just something used on a few short routes/use cases. When designed properly they are far quicker, safer and cheaper even if we had self driving cars on the road.

And it is ludicrous to imagine self driving cars moving around at that speed: (a) there will always be non self driving cars which will cause accidents and (b) self driving cars no matter how good will still cause accidents e.g. sensor failure, poor weather condition, inability to determine future road quality, edge conditions etc.

Why can't self driving cars be public transport? Why will there always be non self driving cars - especially 20 years out? If they're inherently riskier and cause minor irritations (slow down traffic) to major irritations (human driven cars are 6x more likely to cause a human death than a self driving car etc) then how long will we the "i'm in a hurry" allow the indulgence? And interstates are multiple lanes so 2 dedicated lanes to self driving vs 1 or 2 to human driven while we transition.

> EDIT: I do think we'll solve for 200 - 300 mph in 20 - 25 years. But insert 150 - 200 mph.

Honestly, averaging 150-200mph is still hopeless for conventional infrastructure and cars. Hell, 100mph averages are probably out for conventional cars due to fuel efficiency. You could imagine an improvement from co-ordinated drafting (don't try this at home) but that's too marginal to make up the difference.

I'd love to hear your reasoning behind thinking that we'll solve 200-300mph in the next three decades on conventional roads. What kind of vehical are you thinking of?

I think they're gas/electric or all electric motors. The next gen GTR is twin V6 + electric and has top speed of 197mph. Think of how far a Tesla could get 10 years ago vs today too.

For that matter, so long as we're having blue sky thinking involving untold billions of dollars of infrastructure, you might as well bring in new airports well out of the urban centers because that's what hyperloop is likely to be anyway.

Self-driving cars going at 200-300mph for long periods of time seems unlikely. At that point you might as well build lanes for maglev trains.

Also, the hyperloop technology could at the very least be useful for transporting stuff across the ocean (and hopefully cheaper than it is to do it by plane).

Reply on news.ycombinator.com