Saturday, July 11, 2009

41> The Future Revealed!


I have been slipping into the fringe… That’s right, out of the main stream, into the ranks of the eccentric, into the world of the worshipers of the untried and untested… Beware!
I am thinking the unorthodox, have thought the unorthodox, and am beginning to believe! (Scary music and goose bumps ensue)
“What is this unholy philosophy?” you ask…trembling..
“It is this.”
“PRT can and should behave like ordinary traffic. Like ordinary drivers. It should react to the open road like a kid on spring break, unless, of course, Grandma’s on board. Slow traffic should stick to the slow lanes, stay off the highways.”
“No, wait ... Not independent drivers…”(the author closes his eyes as if communing with the great beyond) “Starlings! Schools of fish! Yes, That’s it! Traffic should behave communally. Like drivers who can read each other’s minds, and create ad hoc teams to punch through traffic, like a bubble in a witches brew…”
Think about it. Right now, the prevailing methodology is to slow the whole system down so that a vulnerable few don’t get frightened or motion sick. Everyone else has to take a little longer to get there. I understand that safety is a huge issue. Yet millions are being spent to research ordinary cars “driving by wire” as it is sometimes called. Now I might be a little crazy, but that’s nuts. Speed and switching control while captive in a rail is one thing, but turning your car over, on a real, 3D freeway to a computer is something way higher on the “crazy scale”. Or perhaps not… Modern passenger planes can take off and land unassisted, so what is the big deal?
The possibilities and ramifications are many to this philosophy. For example, passengers can be screened for speed preferences before boarding, or can have their personal preferences set upon creating an account. Now say, for example, 1st and 2nd avenues run North-South, and there are multiple PRT vehicles approaching the area, seeking North-South passage. Why not create a temporary fast and slow lane to accommodate those passenger preferences? Let the speed demons take 1st and the slowpokes take 2nd. This particular arrangement can dissolve as fast as it was created, while circumstances create other opportunities to snake through congested areas.
This obviously takes a much, much higher order of complexity in software control than the ordinary systems that have been developed so far. It is equally obvious that this approach is the future of PRT control, enabling the most efficient use of any given track infrastructure.
The author opens his eyes, séance complete, future revealed! (Wink wink.) You heard it here first, folks… ;o)
All kidding aside, there has been a shift away from centralized control in a variety of communication and control architectures in recent years, as individual “nodes” become endowed with greater and greater processing power. As for PRT, the original concept was to have a fleet of vehicles moving at exactly the same speed. Merges within congested areas would be accomplished by only slight speed changes, or presumably the merge would be called off. The claim of “non-stop” travel was a bit disingenuous, because the trip was to only begin once there was space on the track. In other words, the waiting would be done at the station instead of on route. Now the question is how, with every PRT “pod” having redundant Pentium class processors, can that computing power be employed to prevent traffic congestion in the first place? Although this may all seem very far down the road, I think it does have some bearing on the physical systems, including optimal track layout, station design, and vehicle capabilities.

6 comments:

Anonymous said...

Not sure what you mean by "PRT should behave like normal traffic" but I assume from context that you mean "higher speed when track sections are vacant".

This may sound like a good idea, but factors like drag, energy consumption & cornering forces go up as the square of velocity so you'll see an increase in operating costs (and maybe a LARGE increase) and have to engineer stronger guideway throughout the system.

The need for higher speeds arises when traveling longer distances, so this problem might be solved by providing "express guideways" at intervals & routing appropriately.

In this way you can provide higher speeds when required without the whole system being designed to to meet high speed requirements.

Bryan Williams said...

I think having a fast and slow track is a great idea in an advanced system, well down the road. It does add a level of complexity to the system as well as additional costs. I would say lets get a few up and going first and see how it is received. Personally I'm all for a stronger and faster system. I would much rather arrive quicker to my destination, but I do understand there are people who want to slow down and enjoy the scenery.

Dan said...

Dan the Blogger Responds-

Anonymous, Bryan thanks for the comments.

“Behaving like a driver” mostly refers to autonomous control, as opposed to being directed by a central computer, along with a dash of artificial intelligence.

Let’s take the example of a PRT vehicle transporting an elderly disabled person, and lets assume that the system also includes lighter, faster, two person vehicles. (These would not stress the track any more than larger “family sized” vehicles, in spite of their speed.) Let us further assume that there is a parallel “slow down and acceleration lane” associated with stations, as with most current PRT designs. The slower, larger vehicle, aware that there are 4 or 5 vehicles tailing it, all with occupants who are registered speed lovers, could get off of the main track to let the faster vehicles pass. This is a case where traffic throughput increases by temporarily slowing a vehicle.

I gave the example of an ad hoc fast lane in my post. It is also the case that an ad hoc slow lane was equally created. Other examples: As vehicles leave the station, departure/merge timing could also be influenced by the profile of upcoming traffic. Large numbers of fast or slow vehicles could create convoys, where unlike vehicles are alerted to take alternative routing. If there are vehicles taking cargo only, they can be low on the pecking order. There could even be more expensive “priority” service available, for that matter.

Actually, nearly every convention utilized in ordinary traffic management is an attempt to overcome the limitations posed by roads. (In other words, if everyone had their own road, there would be no need for rules) It would seem to follow, therefore, that in a world of limited track and smart PRT vehicles, similar protocols would make sense.

Bengt Gustafsson said...

I fully agree. There will be a great many opportunities to improve the protocols and procedures to increase capacity, mean speed, comfort etc. When PRT grows a big market there will be a lot of expense and thinking devoted to optimizations, just as in any system. For now we need something simple and easy to understand, that we can present to the public, investors and politicians.

One improvement suggested by Ingmar Andreasson is to increase capacity by platooning empty vehicles. This saves 10-20% capacity for better use. The same idea could be used for cargo of course.

Dan said...

Dan the Blogger responds- I hear you all loud and clear about not getting ahead of ourselves. I just want track and network designs that won't become obsolete, so it pays to look well ahead. Bengt, I looked into that paper by Andreasson and found it most interesting. This is the address for those interested -
http://faculty.washington.edu/jbs/itrans/debate.htm

Unknown said...

We did all that comfort thing with roads and came up with .05g as the limit for comfort. But in a hurry, I once became airborn (1.0g). On planes they give you a paper bag. Or take a pill.
So make it possible to speed.