Talk:Personal rapid transit/discussion

From Wikipedia, the free encyclopedia

[edit] More Real world skepticism of Skytran

Is Wikipedia an encyclopedia or an echo chamber for every ridiculous idea that's techno-geeky and "edgy"?

Just read the silly claims for Skytran on their 2 websites. The folks on this blog[1] point out a few obvious problems with this ridiculous hoax:

"I could write a large book. But I won't. Consider a local area power failure. Consider hanging in a pod, in the winter, untill the utility company restored power, or untill a snorkel truck could come by and rescue you. Consider an ice storm, such as we recently had. Consider 1/2" of ice on the maglev bearing surfaces. Pollution free? This is a NIMBY system. The power has to be produced somewhere. Threquired to move X pounds Y miles weather they come from a big power station, or from burning fuel locally. Imagine buying a weeks worth of groceries for a family of 4. Imagine unloading them at the dismount point a mere 1/2 block from your front door. Existing infrastructure must be retained. An 80,000 lb diesel truck full of bulk cargo will not be replaced by 160 pods. Imagine a bulk shipment to a furnature company. Your new sofa will be delivered in a truck. Instalation: Someone with a copy of Photoshop(r), please take some of the pictures on the 'Photo tour', and put them a more typical urban background. Use narrow sidewalks and roads typical of an urban environment. Include some colorful grafitti on the pods. Cost: This is new technology. It's possible to build known technology (roads or rails) pretty close to estimates ( 8) This is not to say that it can't be made to work. It's just that I expect that it won't work as advertised, or anywhere near within budget. The energy that is currently self contained in motor vehicles, will be demanded from the public power grid. I don't know how things are in your neck of the woods, but we don't have that much surplus capacity, and trying to add it would bring howls of protest from the very 'Greens' that would demand that SkyTran be implemented."

I'd be interested in seeing how they plan to switch these cars from one track to another. Monorails are notoriously difficult to design switches for, and I can't think of any that currently exist with high-speed switches. Something in an urban area will require LOTS of them, and I'll get switches will provide the bulk of the infrastructure cost.

This will definitly not be as convenient as cars. I can currently get in my truck in my garage, drive to work, and get out in the garage at work without being rained or snowed on. Who wants to walk a couple hundred yars to get to the on/off loading points?

And imagine how nasty the interiors of those cars are going to get. All those teenagers going on dates, riding home drunk, realizing this is their last opportunity alone together before getting back to the parentals. You want to ride in that car afterwards?

Administrator- Please delete the Skytran page and delete all PRT pages and replace them with a page that explains that PRT is a hoax...which it is.Avidor 12:40, 14 April 2006 (UTC)

What about the robot? That's my favorite part. -- Transit Guest 16:07, 14 April 2006 (UTC)

Heh, the robot's a little ridiculous. I'm sure it could be designed to work *profitably* - but it won't. That would take too much time and money. Anyway, I only skimmed your comment, but I found two things i have answers to. Most PRT proposals propose "passive switching" or "vehical switching" in which the switching mechanim is in the vehical rather than the track. This is a-kin to a car turning ("switching") from one lane to another. Also, I *think* (tho i'm not sure - and don't use that as another quote for an ad hoc attack) that most PRT systems proposes self powered vehicals, I know that SkyTran does, TriTrack does, ULTra does, etc. Thus one wouldn't get stuck in a power outage. It might be a good idea to note these generalities in the main article. Fresheneesz 05:59, 15 April 2006 (UTC)

First, that was Avidor's long comment.

Second, if the robot takes too much money to develop, then by definition it's not profitable - regardless of the time.

Third, as for switches - yes, the on-board switches (within the enclosed rail) make sense. But if the cars are all self-powered, when do they refuel/recharge? I thought the point was that they were electrical, running on the grid, like a third-rail system or something? Because batteries are REALLY heavy - that's what makes fuel-cell hybrid autmobiles so much more attractive than full-electric battery cars, after all. -- Transit Guest 13:59, 17 April 2006 (UTC)

If the cars recharge at stations, they won't need to go very far on each battery charge. However, I also imagine getting electricity off the track would be cheaper. Stephen B Streater 14:06, 17 April 2006

"if the robot takes too much money to develop, then by definition it's not profitable" - obviously I defined "too much" differently from you. I meant it akin to this: "a farmer selling his years crop of 20 apples (say), can sell those apples for $1 each, but if he bough $10,000 of land, and $90,000 to research, develop, and build the machines to run it, then he could sell 20,000 apples for 99 cents each." - obviously this is a very rough exageration, but I meant that the farmer won't go to such trouble to do this - even tho it would pay for itself within 6 years.

Also, SkyTran does not plan on running electricity on a grid, because of the expense (building it into the track, and maintenance on both track and vehical). I'm not sure i'm convinced whether grid powered or battery powered would be more economical. Fresheneesz 18:39, 17 April 2006 (UTC)

[edit] Discussion of Robot Construction of Skytran

Picture of the guideway-extruding ROBOT [2]

Okay, is that enough evidence declare PRT a total joke? Why aren't the other so-called PRT companies condemning Skytran for being such a joke and giving the rest of them a bad name?

You can't get sillier than this statement "The light weight per foot of the track design also allows the use of a semi-automated track forming manufacturing robot (much simpler than the Robosaurus machine)."

Here's the Robosaurus[3]

Administrator, it's time to declare this and other PRT pages a bad joke and delete it. Avidor 19:19, 14 April 2006 (UTC)

Oh? Whos calling for censorship now? Malawiki has lots of experience with heavy machinery and strong machines. Note also that this machine is not planned to move at break neck speeds, it moves at a mile per DAY. This is so the track does not snap from strain. In any case, I noted my opinion of this idea in the previous section. Avidor, this article will not be deleted - quit trying. Fresheneesz 06:02, 15 April 2006 (UTC)

The chances of a machine such as that being used to form the track are remote at best. But since there is no such thing as Skytran, it's irrelevant to the article anyway. Just zis Guy you know? 07:19, 15 April 2006 (UTC)

"a semi-automated track forming manufacturing robot"

Doesn't "semi-automated" mean that human operators are involved in running the machine? Doesn't this just make the thing a task-specific piece of equipment? What annoys you is the word "robot," isn't it?

It's one man's concept. If it's not feasible it means it would have to be manufactured and assembled more conventionally, bringing installation costs in line with ULTra, Vectus or Skyweb. Why not move on to a discussion of the current state of the art of permanent magnets, and whether those are sufficiently advanced to levitate small light vehicles? -15apr06 1:33 CDT

Robot-haters may be interested in this description from one of the Seattle newspapers, of their light rail tunnel boring machine:

"Three months into the job, you might say that Sound Transit's new Beacon Hill Tunnel has been simply boring...

The one-mile Beacon Hill segment alone is costing $292 million — a $72 million increase from what was estimated three years ago. ...Grout had to be injected into the hill, so water and loose sand don't seep into the underground station, where an elevator shaft for future passengers is being mined 16 stories down from the hilltop.

...Like a rotary razor, the blades on a 21-foot drill head churn at a typical speed of 1 ½ revolutions an hour. Dirt drops toward a rotating screw that pushes it out the back end.

To prevent cave-ins, Mitsubishi of Japan built the drill so it can also install the concrete tunnel wall, right away. At the back end, a robotic arm snatches a concrete, circular section of the wall — like picking up a six-pack of beer with two fingers, one tunnel manager said — and sets them in place. Workers bolt the curved pieces together, to complete a solid ring.

Hmmm, sounds like a semi-automated task-specific piece of equipment, doesn't it? -21apr06, 1:55p CDT

Please note that my final year degree project was in robotics and my first two jobs after graduation were both in industrial automation, latterly writing real-time control software. I still think the whole thing is a fantasy. Just zis Guy you know? 19:01, 21 April 2006 (UTC)

The operative word here is "think". Apparently, if you are an admin, all you need to do is "think" something is fantasy and you can start whacking articles off the map. Like Avidor, JzG refuses to refute the technology with scientific or engineering arguments, instead resorting to the "this can't possibly work" line of reasoning. It's basically a technophobic position that ignores the realities of modern technology and the fact that systems much more complex than PRT have been designed and built. PRT's problems are not technological, but political. Unfortunately, as a respected admin, JzG's opinions are treated as fact by many Wikipedians. A Transportation Enthusiast 21:43, 21 April 2006 (UTC)

Any big engineering project will encounter technical difficulties. This is one reason for the political difficulties. The onus is on the proponents to convince the sceptics. When a system is built, the sceptics here will be convinced. Stephen B Streater 21:56, 21 April 2006 (UTC)

The point of the WP policy of reporting what reliable sources say about things is that the editors of WP don't have to be experts themselves for this to work. If a statement in the article requires original research or a deep first hand understanding of the issues to the extent that significant proportion of editors don't get it, people should relax if it doesn't get it into WP. Stephen B Streater 21:56, 21 April 2006 (UTC)

My feeling is that the article should be not strident or aggressive about what is definitely possible for things which haven't been financed and built. A lighter touch now can be strengthened as technology improves (as it always does) and systems start to arrive. Stephen B Streater 21:56, 21 April 2006 (UTC)

[edit] Braking distance

The issue of braking distance interests me. Carriages in a train have effectively no braking distance between them, and this is deemed OK. When pods are moving very close together, they must also be moving at very low relative speed, making the system safer - it effectively turns into a train. As someone who drives a fly by wire car, I don't have a problem with close distances. Stephen B Streater 05:57, 29 April 2006 (UTC)

You have a fly by wire car? Does that mean it doesn't have a steering column? Anyway, I suppose the reason that train cars having negligible spacing is because the train is considered as a whole. After all, noone worries about the spacing between two seats in a train car, or between the seat and the person sitting in it. There have been PRT proposals that propose systems without spacing between vehicals. But since PRT vehicals aren't physically connected to eachother, I can definately see the concern.

Actually, it's all fly-by-wire except the steering. Mostly braking and acceleration and CVT gear ratios. See Prius. Stephen B Streater 14:24, 30 April 2006 (UTC)

For example, in an emergency braking situation where perhaps 3 or more g's are pulled, if another vehical was 1 meter behind the braking vehical, they would colide at about 7.5 m/s . This is of course assuming that the rear vehical wouldn't have had time to respond with its own emergency brakes in the 1/4 second it had. Far spacing and close spacing isn't the concern - what really matters are those critical distances where relatively high speed colisions become possible. On the other hand, such a colision would probably not be very harmful to passengers, although the vehical itself might be done for.

I think 15m^-2 is a reasonable speed too. The cars could know within microseconds that something was up. They could even coordinate braking by giving an absolute time with their synchronised clocks saying brake at 12:32:34.102034 pm at 3G. It would feel that they were connected into a train.

After all that, the bottom line is if regulatory agencies would allow it. And since we all know how well the government would score on a 10th grade math test, I doubt there would be much support. Fresheneesz 11:26, 30 April 2006 (UTC)

All it needs is a big private venture to do it where safety is less of an issue - Disney Shanghai perhaps. If Airbus can fly, so can PRT. Stephen B Streater 14:24, 30 April 2006 (UTC)

I agree, a big private venture may be the best hope for PRT getting into the market. I know someone has pitched the idea to Microsoft for their sprawling campus, though I don't know what the status of that proposal is. Disney might be the perfect company to do it with their amusement parks, especially in Florida where they have a real long-range transportation need. But then, if Disney was the first to build PRT, detractors would be the first to claim that it's nothing but a toy, a Disney "ride" -- fine for a cute attraction but not applicable for the "real world". So a Disney PRT, even if successful, might be a double-edged sword for PRT. A Transportation Enthusiast 17:36, 4 May 2006 (UTC)

There are several points on the whole headway/braking issue that I've had to work through in my head. I initially was skeptical of short headways until I really thought about it. Here are my insights on the matter:

• If you have a 3-car train, the trains are physically connected. Therefore, when the lead car stops, there is a physical constraint that causes the following vehicles to also stop. PRT has no such physical linkage, and there seems to be a basic distrust, among skeptics and regulators, of the capability of PRT controls to emulate that physical connection. In other words, PRT opponents seem skeptical that a virtual linkage between vehicles is as reliable as a physical linkage.
  
  Now, it's true that it's much easier to design a steel connector than it is to design a complex multi-vehicle control system, but that doesn't mean that such a control system cannot be built to be as reliable as that steel connector. It's just a matter of good engineering. And some systems have actually been successfully tested to prove that it's possible (CVS, Cabintaxi). Of course, there's also the French system, which failed miserably, but that only proves that it's possible to do it badly. :-)

• There seem to be two approaches to hard braking in PRT: (1) SkyTran: provide automobile-style vehicle passenger restraints to allow for the occasional hard braking scenario (i.e. 0.5g all the way up to 6g for SkyTran), and (2) most other systems: design the system so that hard braking is so rare as to be inconsequential -- in other words, only a very rare sequence of events coupled with a simultaneous act of God will cause a hard-braking scenario.
  
  I prefer option (2), since it focuses on preventing failures from occurring in the first place. I think Taxi 2000 was designed for maximum emergency braking of 0.4g, even at low headways. Harder braking was all but eliminated by fail-safe design using checked redundant systems for all crucial components. Of course, Taxi 2000 systems were never built and tested, so all we have is the engineering. I would have loved to see a full scale Taxi 2000 prototype built to test out their engineering, but it seems their window of opportunity may have passed.

• There is also the infamous "brick wall stop" issue. This has been the main barrier to sub-2-second headways in PRT systems. But I've thought at length about this issue, and on a dedicated guideway, the only way to have a brick wall stop scenario is (a) a vehicle derailment or (b) an act of God (i.e. sink-hole swallows up a section of guideway). From what I've seen, case (a) can be designed around. Systems supposedly have been designed so that derailments are extremely rare (again, much of this is pure engineering, untested).
  
  The main issue seems to be during switching -- obviously a failure during a switch could theoretically cause a vehicle to slam into a diverge point. So there has to be careful design and sufficient redundancy to eliminate this scenario. But other than a vehicle derailment as described above, typical failures will not cause a brick wall stop, but rather a slow descent to stop. This is a scenario that can easily be dealt with, without injury. As for (b) - acts of God - no system can guard against this. I'm convinced that, with proper design and engineering, brick wall stops are not an issue with PRT. But rail experts would tend to disagree, since their experience is with fundamentally different system. And so the debate rages.

• PRT is actually less susceptible to brick-wall stop scenarios than most other transit systems, i.e. at-grade light rail and buses. Any system that doesn't have a dedicated ROW is going to be more susceptible to collisions with other vehicles (cars, trucks, even other trains/buses). And yet, even then, collisions are relatively rare. I can't believe that a properly designed PRT system would ever be susceptible to brick-wall stopping scenarios. Key words here: properly designed.

• This is the big point for me: let's say the incredible happens and there's a brick-wall stop on a PRT guideway running at half-second headway. Tragically, the occupants in the first vehicle perish immediately, along with probably the next 5 or 6 vehicles - for a total of, maybe, 10 fatalities (given ~1.5 persons/vehicle). Vehicles beyond the 6th vehicle can stop in time to avoid killing its occupants, so they're all safe.
  
  Now, this is an absolute worst case disaster for PRT, probably involving an act of God, and only 10 people have perished. Now consider a bus or a train running full speed into a tractor trailer fully loaded, causing a near brick-wall stop in that train or bus. This is theoretically a more likely scenario than the PRT stop, since trains and buses cross intersections. In the case of a train, the vehicles probably derail and cause perhaps dozens or even hundreds of fatalities. In a bus, it's probably on the order of a few dozen. But in either case, the fatality count is much higher than the equivalent PRT disaster.
  
  The basic fact is, the low-density nature of passengers in PRT means that a disaster causes fewer fatalities than the equivalent incident in a packed bus or train. And this, even at half second headways. Couple this with the fact that PRT's dedicated guideway would (theoretically) make such disasters extremely rare, and PRT, even at low headway operation, could be perhaps an order of magnitude safer than even safe transit modes. But again, this is all speculation, because PRT systems have no safety data yet. Despite this, it seems quite logical to me that a separated guideway will have fewer brick wall stops than an at-grade guideway, and that highly dense passenger systems will have more injuries/fatalities per incident than lower density passenger systems.

Anyway, those are my thoughts on braking and headways... A Transportation Enthusiast 04:43, 3 May 2006 (UTC)

I hope you don't mind ATE - I broke your response into more managable paragraphs. "I prefer option (2)" - I just want to point out that option "1" and "2" aren't mutually exclusive - and SkyTran focuses on both issues, option 2 is basically inherent to PRT.

"a failure during a switch could theoretically cause a vehicle to slam into a diverge point" - I've never actually thought of that scenario, but its a good point. Is there any way you guys can think of to design a diverge-point to counteract wedging? I can imagine, if the tip of the diverge point and the tip of the vehical's point of first potential contact were both pointed, then the changes of a brick-wall stop would essentially be 0. Fresheneesz 05:30, 4 May 2006 (UTC)

No problem on the paragraph breaks... I actually didn't break the points into paragraphs because of the bullet indenting -- I don't like the way that a new paragraph within a bullet-item list loses the bullet-indentation. But I think I've found a workaround that seems to do the trick: I inserted <br><br> to add newlines without starting a new paragraph, and the indenting is preserved. This makes me much happier. :-) A Transportation Enthusiast 11:37, 4 May 2006 (UTC)

Re: SkyTran braking - true, SkyTran can be both (1) and (2), but then I would argue that (2) should be sufficient to do the job, and if SkyTran is designed so that hard braking is extremely rare (i.e. only catastrophic failures cause a hard brake) then (1) shouldn't be necessary. The key point here (which I neglected to mention in my original comments) is catastrophic failure. The system must be designed such that only catastrophic scenarios, typically involving an act of God, can result in a hard brake.

So, in essence, if a system is designed for case (2), then (1) becomes superfluous. The analogy I like to use is: we don't put parachutes on passenger jets, because they add significant overhead and are only (theoretically) useful in a catastrophic failure. Ultimately, the reality is, no amount of human engineering can completely protect us from acts of God. A Transportation Enthusiast 11:59, 4 May 2006 (UTC)

Re: diverge point failures - I believe this is a tricky problem, but one that has largely been solved in most system designs. I don't know the exact approach they use (and they may vary across systems), but I believe it has something to do with applying the switch and verifying that the switch occurred long enough before the intersection so that the vehicles can be braked in time in the event of a switch failure. I also believe that the switches themselves are checked-redundant, so outright switch failures are already pretty rare.

Think of it this way: the window of catastrophic failure for a switch is the time after the switch has been applied and verified, but before the intersection is reached. If the switch somehow fails during this short interval, then you might have a brick wall stop. But within this short window, you'd have to have simultaneous failures of at least two redundant components, in such a way as to create the wedge scenario. The task of the PRT system designer is to ensure that the probability of such a simultaneous failure is extremely (catastrophically) rare. (For example, a bolt of lightning just at the right time and place could theoretically cause such a failure -- but as I said, there's only so much that we can do to protect against acts of God...) A Transportation Enthusiast 12:13, 4 May 2006 (UTC)

Well, disregarding god, there are certain small probabilities of "catastrophic failures". While the probability is so low that riders can feel "certain" that they're safe, such failures are bound to occur. Protecting against brick-wall stops has a couple upsides. The downside is obviously the extra cost of an emergency brake and a seatbelt (in the case of SkyTran). The upsides are that 1) in the case of a brick-wall stop, less people are likely to die, 2) its a plus for safety in the eyes of regulatory agencies 3) something like a seatbelt would prevent other random injuries from happening at moderate braking speeds (not extreme emergency, but not normal). I don't know about you, but if I were a regulatory agency, I wouldn't let PRT go online without emergency brakes. And then seatbelts aren't a huge expense anyway.

Applying the switch action long before the actual switch sounds like a good idea. I'd feel better if my idea were also designed : ) . Remember, though, that engineers aren't perfect and things that look like acts of god might actually just be small engineering errors. And thats what redundancies like emergency brakes are for. Fresheneesz 20:21, 4 May 2006 (UTC)

"such failures are bound to occur" - not necessarily. Components can be designed such that MTBF is literally millions of years. In that case, the overwhelming probability is nothing will occur for the the life of the component, under normal operating conditions (i.e. no act of God). So, cars have seatbelts and airbags because there is a relatively high probability of high-impact collisions in cars. But LRT trains and buses do not have such restraints, because the probability of an auto-type collision, though not precisely zero, is so close to zero that the restraints are deemed unnecessary. I think properly-designed PRT would fall into this category.

There's another consideration: putting seatbelts in PRT vehicles may give the false impression that the passenger is in for a wild ride. It might also cause some people to believe that PRT is somehow less safe than other modes like trains and buses, which do not "require" seatbelts. For something as politically charged as PRT, that might not be a message that you want to send if you are a PRT designer. A Transportation Enthusiast 20:20, 5 May 2006 (UTC)

Magnetically levitated vehicles wouldn't need switches on the track, would they? Couldn't you have the switches on the vehicles, with a failure mode of going straight into the "fix me" depot? Stephen B Streater 21:12, 4 May 2006 (UTC)

I think that most PRT systems have the switch on the vehicle itself. Everything I wrote about above assumes vehicle-based switches, not track-based switches. A Transportation Enthusiast 20:20, 5 May 2006 (UTC)

Right, when we were talking about switches we meant on-vehical switching mechanisms. I think a general consensus is that mechanical on-track switches make PRT unworkable, because of the excessive wear and tear those switches would undergo.

"designed such that MTBF is literally millions of years" - MTBF for components is much different than MTBF of a system. Any given component might have a MTBF of millions of year, but as a system, those failure rates are compounded. Not only that, but larger constructions made of smaller components will have their own (earlier) MTBF rate.

"LRT trains and buses do not have such restraints, because the probability of an auto-type collision ... is close to zero" - This is not true in a large sense. Think about the weight of LRT trains and busses. Those things don't slow down for any old Hum Vee that decides to get in its way. Bus crashes may be less than normal car crashes (or might not, I don't know), but the *real* danger is the rate at which deceleration occurs in a crash. Busses don't slow very quickly, therefore any sort of accident is probably going to be less lethal than a car version. Fresheneesz 07:45, 6 May 2006 (UTC)

You are correct, the MTBF for systems is compounded by the MTBF of its individual components. But only certain select components can fail in such a way as to cause a brick wall stop. It is my understanding that the aggregate MTBF of those critical components can be made high enough so that it is not an issue.

Re: bus decelerations - in a bus-car collision, the bus is much more massive than the car, and the bus deceleration is not severe enough to cause significant injuries for the bus riders. But this is not necessarily true of bus-bus collisions, or bus-train collisions, or bus-trailer collisions, or bus-tree collisions. All of these are very rare, but not impossible, scenarios, and they will cause potentially fatal decelerations for the passengers in the bus. Same with trains: they are very safe, especially compared to automobiles, but they do crash and derail once in a while. Despite this fact, buses and trains do not have passenger restraints, because the possibility of such a crash is so small that restraints are deemed unnecessary. The same can be said of properly-designed PRT, I believe. A Transportation Enthusiast 14:32, 7 May 2006 (UTC)

Yea, combined with the falacious psychological "seatbelts mean its less safe" issue, its probably a good argument for no-seatbelt. However, if regulatory agencies do enforce long headway restrictions, then a seatbelt might be the only way to allow the high G stops neccessary to make 1 second headways a reality. If the choice is between 5 second headways without seatbelts, and 1 second headways with - I think its more cost effective to have the seatbelts (in terms of the whole system's effects). Fresheneesz 03:11, 8 May 2006 (UTC)

Agreed. But it would be a political decision, not a technical one. Perhaps ULTra's experience at 2 seconds (and no restraints) will change the political environment for the better... A Transportation Enthusiast 06:23, 8 May 2006 (UTC)