Hacker News Reader

Comment by Stevvo | Hacker News Reader

Are you claiming that trains running outside cities are less suitable for one person operation? There is no evidence of that, plenty of trains around the world run both inside and outside cities with a single operator. There is also no evidence of a city being a "more predictable environment". Deaths are roughly equivalent; in the city it's people jumping on tracks intentionally, outside its drivers getting stuck on crossings.

View on HN

There are many other problems.

A subway or city train stops every few minutes. This means that if somebody gets hurt, has a stroke, assaults somebody, starts shooting up… there is almost immediately a way to board more staff and handle the situation.

On a train with hours and tens or hundreds of kilometers between stops this is much less the case.

I mean things like: beyond city scale, you can't assume all the technology has been tested together, you can't assume you actually know where other trains are, etc.

You can't assume those things in a subway, either. GPS doesn't work underground.

The trains ride on a track. The tracks have sensors. The sensor positions are known. Now you know where all of the trains are. GPS satellites could drop out of the sky and you’d still know where the trains are.

Yes, and those "sensors" have very limited precision. There's only one every several hundred yards.

>Yes, and those "sensors" have very limited precision. There's only one every several hundred yards.

What?

There’s one as far apart as you put them. The precision is what you make it.

And every sensor you add is additional cost, not just upfront, but maintenance for ever.

And not having sensors results in not knowing where the trains might be on your track _forever_.

I’m not entirely sure what you are trying to argue. Your original comment said you can’t know where trains are on underground subways because GPS doesn’t work underground, but that’s not the technology they use anyways, so it doesn’t make sense to reference it. The you seemed to argue that what might currently be in place is insufficient, but that wasn’t what this thread was talking about.

You keep putting words in my mouth. I am just pointing out the usual HN tech dork ignorance of how real world mechanical systems actually operate. It's like any time an aviation thread is around here, the rank ignorance is incredibly painful.

I haven’t put any words in your mouth, just summarized what I read.

The only one showing ignorance here seems to be you suggesting subways use GPS.

The latest tech for CBTC uses UWB for train positioning, which has a precision of a few inches.

New York did some trials with UWB several years ago, which were claimed to have been successful. Since then, there has been very little news of permanent deployment or future developments either in New York or elsewhere.

Not sure whether this means that the various signalling suppliers are waiting for a customer to take on the technical risk of being the first to deploy UWB CBTC, or that there were costs or disadvantages that make it unattractive.

I work in rail. In any automated system, you don't assume anything; you either know it absolutely (fail-safe systems), or you revert to pure manual control at line-of-sight of the operator, which is far slower.

And yes, all the technology has been tested together. Before a train can enter a new region of authority, it must have permission from that authority, which means they must prove that the train can work with the local tech.

Reply on news.ycombinator.com