>i know what you mean. losing webvan was a
>bummer. in SF, you can do pretty well using a
>combination of the BMW (meaning BART + MUNI +
>Walk) plus CityCarShare (www.citycarshare.org)
>for doing things like big grocery runs or
>hardware store trips.
PRT has not been created to totally replace the automobile.
PRT has been designed as public transit, its goal is to reduce traffic congestion by attracting more riders than conventional transit. So the question shouldn't be "how do I carry 15 bags of groceries on PRT?", but "do I carry 15 bags of groceries on the bus or train TODAY?" You don't of course, nor would you on PRT, any more than you would take the bus to pick up a load of lumber at Home Depot.
But you would take PRT to work, or school, to a restaurant, or to the movie theater. Or to get ONE bag of groceries.
The PRT innovation is that it goes station-to-station nonstop. The conventional "mesh" as you describe it requires transferring, which is a recognized disincentive to using mass transit.
Oh. Well... If The French couldn't make it work...
The problem with sweeping statements like "This is an idea that doesn't work" is that it assumes past PRT projects were all textbook cases of proper implementation. The corollary to "no one that I know of has succeeded with this, perhaps they've all failed!" is this hasn't worked yet, therefore it will never work, which is clearly nonsense.
The fact is that modern PRT designs have never been given a fair test. The Boeing and Raytheon PRT projects were not implemented by the people who designed them, and the implementers made serious and avoidable mistakes. Arguably politics led to design compromises, there were no inherent flaws in having small vehicles, small guideways or automated control.
The people who worked on Aramis seem to have been bad at the politics of the French technology sector-- but mostly it appears to have simply been a really bad design. It was probably a matter of timing. Aramis's cancellation in the mid 80s was exactly when the current version of the Taxi 2000 PRT got going. ULTra and Skytran started in the 1990s. All three designs have benefited from 1980s-90s advances in microcomputers, software, propulsion, and lightweight composite construction materials. ULTra benefits from the latest generation of batteries.
Finallly, it's important to note that the designer of ULTra is part of the group that is implementing it. As a result the prototype now being tested appears to be exactly as it was originally designed. ULTra has an excellent chance of succeeding.
Malfunctional: in a well-designed PRT system the key components have redundancy: 2 motors, 2 steering switches, 2 control computers, etc. There are estimates which, taking into account proven component reliability (there are no exotic, futuristic components), have estimated the time between mechanical breakdowns which lead to a vehicle stopped on the guideway as being on the order of MILLIONS of years. As a fallback, some systems provide for a breakdown to be pushed to the barn by another vehicle.
Careen out of control into other cars: PRT cars don't operate independently, but under a "check system" overseen by a central computer. How the different designs accomplish this vary, but the basic idea is this: Each car's computer reports many times per second (in one system it's every 40 milliseconds)-- location, speed, etc. The central computer independently senses the speed, location, etc. of all the vehicles, and then compares the two sets of data. If they don't match (or even if a vehicle reporting signal is interrupted), then that's a Trouble situation and the system tells all the vehicles behind the Trouble Vehicle to stop, and to detour others from entering the section of guideway where the trouble is occuring.
Overriding the system: Programming a PRT car isn't like starting up an automobile. Rides are purchased from an ATM-like machine in the station. Tickets are coded with the selected destination. The rider puts the ticket into a slot on the outside of the vehicle or in a turnstiley box on the boarding platform. This gives the vehicle the destination and opens the door. The navigation systems in the vehicles are not accessible, it would be stupid to design it any other way.
The whole reason for the PRT concept IS the dispersal of housing and business. Right now the only choice transit "experts" will put forward are line-haul systems-- light rail and monorail, which are exactly wrong for dispersed cities.
This is why PRT systems would have their guideway arranged in a grid, not a corridor, not a hub-and-spoke. Lightweight PRT vehicles translate into slimmer guideway constructed from fewer materials than light rail or monorail, therefore cheaper per mile, therefore a city-wide grid is feasible on cost.
What would most people rather do: travel when they need to travel, or when a timetable dictates?
1. Because PRT would operate on-demand, people are being picked up and dropped off continually.
2. In an ideal system, there are enough stations so that everyone is within walking distance (about a quarter mile) of a station.
Therefore, chances of congestion are reduced two ways: removing the need to wait, and reducing the service area radius for each station. Big stations are cool if serving a timetable-dependent form of transit, but they are expensive at about $4 million each, $10-12 million if elevated.
Additional benefits:
1. Because all neighborhoods have a station, no neighborhoods are overrun by traffic of people from adjacent areas coming in to access transit. Costs of traffic mitigation and parking lots/garages are eliminated. Neighborhood character is respected.
2. Social equity: If everyone in a city is paying taxes to build the system, PRT gives them benefit in the form of convenient access. Since PRT is on-demand, each district gets the amount of service the request.
Mr. Malewicki has acknowledged that the philosophical basis of Skytran owes a great deal to Prof. Ed Anderson's work on PRT theory and his Taxi2000 design. A great deal of the Skytran web site reads like Anderson's-- (begin sarcasm) although I'm not charging plagiarism, I'm sure it is an homage. (end sarcasm)
The main problem with Skytran, as I see it, is that their cars are clearly not ADA compliant. It will probably have to be larger, which probably will mean larger guideway to support them and therefore higher costs (still cheaper and less visual impact than conventional monorail though).
The other problem I haven't seen addressed is the millions of passive mag coils which would be packed into the Skytran guideway-- how far out of alignment could one get without affecting the ride?.
The Morgantown system, which is still operating by the way, is called "The PRT", but it is not Personal Rapid Transit. Morgantown is a People Mover-- rides are not non-stop during certain periods, and the cars are large, 20+ persons each. Big vehicle = heavy = huge guideway = more materials = expensive. Thus, the implementers (the US Government) ignored a basic PRT principle: Keep It Small.
The Raytheon PRT project in the 1990s suffered from the same fate. A defense contractor, Raytheon did not seem to think it important to keep costs low-- the vehicle was big, so the guideway was big, etc. They also made a design mistake: the tire diameter was too large for the part of the vehicle that rode inside the guideway, necessitating larger guideway dimensions. However, Raytheon's project was a technical success in that the control software and in-vehicle steering mechanism worked. If they had kept the vehicle size & weight small, the guideway structure would have been cheaper, smaller and less obtrusive, and it would have been a cost success as well.
You're exactly right on the private ride and security features, although I'm not sure if ULTra has a panic button.
There's an additional deterrent: Only the person buying the ticket knows the destination ahead of time. If you were a mugger or rapist, would you force your way onto a monitored Taxi2000 car without knowing where you were going, and the only control you have is to hit the Panic Button, which only means alerting the police? Of course you wouldn't.
You mount cameras in the station and inside each vehicle. Because a smartcard is used to program destinations, you can match the time of vandalism agains the log of smart card usage.
ULTra and Taxi2000 PRT cab drive and control mechanisms are inaccessible to the rider. If a car is unusable, it is sent to the maintenance barn, or if it breaks down it is pushed by another cab.
A PRT station is off-line on a siding, so a car stopped at Station X doesn't delay a car going to Station Y.
Boarding and unboarding doesn't cause congestion in a station because an empty cab is used by the next rider. If someone needs to arrive at a station full of empty cabs, the system knows and clears out a berth. The empty cab goes to a station where it is needed.
Serial vs Parallel: The main reason a subway picks up and drops off a crowd of people at the same time is because the schedule requires people to show up at an appointed time & ride together. PRT is on-demand, people ride when they want. Therefore, no crowds at stations, therefore huge stations aren't needed.
1.
The number of little driverless cabs is less than the number of cars because the little driverless cabs are reused by a succession of riders. Trips by little driverless cabs are shorter than trips by car, bus, or train because little driverless cab trips are nonstop.
Suppose each little driverless cab carries 5 different fares per hour. If a system has a fleet of 500 little driverless cabs, that's 2500 fares per hour. 2000 little driverless cabs = 10,000 per hour.
2.
The idea behind Personal Rapid Transit is to eliminate the need for transfers. There would be lots of tiny stations (smaller than a house) throughout the network, enough so that you would usually be within walking distance of your origin or destination.
3.
PRT systems are envisioned as INTRAcity or city-to-suburb PUBLIC transit. If you need to leave the area or travel 300 miles, then you'd take a car, plane or train. If you need to pick up lumber for a home improvement project, you'd use your SUV (if you have one) or have the lumber yard deliver. PRT is designed to alleviate urban traffic congestion by increasing transit's share of trips. PRT does not envision totally replacing cars.
Slashdot Mirror
>i know what you mean. losing webvan was a
>bummer. in SF, you can do pretty well using a
>combination of the BMW (meaning BART + MUNI +
>Walk) plus CityCarShare (www.citycarshare.org)
>for doing things like big grocery runs or
>hardware store trips.
PRT has not been created to totally replace the automobile.
PRT has been designed as public transit, its goal is to reduce traffic congestion by attracting more riders than conventional transit. So the question shouldn't be "how do I carry 15 bags of groceries on PRT?", but "do I carry 15 bags of groceries on the bus or train TODAY?" You don't of course, nor would you on PRT, any more than you would take the bus to pick up a load of lumber at Home Depot.
But you would take PRT to work, or school, to a restaurant, or to the movie theater. Or to get ONE bag of groceries.
The PRT innovation is that it goes station-to-station nonstop. The conventional "mesh" as you describe it requires transferring, which is a recognized disincentive to using mass transit.
The problem with sweeping statements like "This is an idea that doesn't work" is that it assumes past PRT projects were all textbook cases of proper implementation. The corollary to "no one that I know of has succeeded with this, perhaps they've all failed!" is this hasn't worked yet, therefore it will never work , which is clearly nonsense.
The fact is that modern PRT designs have never been given a fair test. The Boeing and Raytheon PRT projects were not implemented by the people who designed them, and the implementers made serious and avoidable mistakes. Arguably politics led to design compromises, there were no inherent flaws in having small vehicles, small guideways or automated control.
The people who worked on Aramis seem to have been bad at the politics of the French technology sector-- but mostly it appears to have simply been a really bad design. It was probably a matter of timing. Aramis's cancellation in the mid 80s was exactly when the current version of the Taxi 2000 PRT got going. ULTra and Skytran started in the 1990s. All three designs have benefited from 1980s-90s advances in microcomputers, software, propulsion, and lightweight composite construction materials. ULTra benefits from the latest generation of batteries.
Finallly, it's important to note that the designer of ULTra is part of the group that is implementing it. As a result the prototype now being tested appears to be exactly as it was originally designed. ULTra has an excellent chance of succeeding.
Careen out of control into other cars: PRT cars don't operate independently, but under a "check system" overseen by a central computer. How the different designs accomplish this vary, but the basic idea is this: Each car's computer reports many times per second (in one system it's every 40 milliseconds)-- location, speed, etc. The central computer independently senses the speed, location, etc. of all the vehicles, and then compares the two sets of data. If they don't match (or even if a vehicle reporting signal is interrupted), then that's a Trouble situation and the system tells all the vehicles behind the Trouble Vehicle to stop, and to detour others from entering the section of guideway where the trouble is occuring.
Overriding the system: Programming a PRT car isn't like starting up an automobile. Rides are purchased from an ATM-like machine in the station. Tickets are coded with the selected destination. The rider puts the ticket into a slot on the outside of the vehicle or in a turnstiley box on the boarding platform. This gives the vehicle the destination and opens the door. The navigation systems in the vehicles are not accessible, it would be stupid to design it any other way.
This is why PRT systems would have their guideway arranged in a grid, not a corridor, not a hub-and-spoke. Lightweight PRT vehicles translate into slimmer guideway constructed from fewer materials than light rail or monorail, therefore cheaper per mile, therefore a city-wide grid is feasible on cost.
What would most people rather do: travel when they need to travel, or when a timetable dictates?
1. Because PRT would operate on-demand, people are being picked up and dropped off continually.
2. In an ideal system, there are enough stations so that everyone is within walking distance (about a quarter mile) of a station.
Therefore, chances of congestion are reduced two ways: removing the need to wait, and reducing the service area radius for each station. Big stations are cool if serving a timetable-dependent form of transit, but they are expensive at about $4 million each, $10-12 million if elevated.
Additional benefits:
1. Because all neighborhoods have a station, no neighborhoods are overrun by traffic of people from adjacent areas coming in to access transit. Costs of traffic mitigation and parking lots/garages are eliminated. Neighborhood character is respected.
2. Social equity: If everyone in a city is paying taxes to build the system, PRT gives them benefit in the form of convenient access. Since PRT is on-demand, each district gets the amount of service the request.
The main problem with Skytran, as I see it, is that their cars are clearly not ADA compliant. It will probably have to be larger, which probably will mean larger guideway to support them and therefore higher costs (still cheaper and less visual impact than conventional monorail though).
The other problem I haven't seen addressed is the millions of passive mag coils which would be packed into the Skytran guideway-- how far out of alignment could one get without affecting the ride?.
The Raytheon PRT project in the 1990s suffered from the same fate. A defense contractor, Raytheon did not seem to think it important to keep costs low-- the vehicle was big, so the guideway was big, etc. They also made a design mistake: the tire diameter was too large for the part of the vehicle that rode inside the guideway, necessitating larger guideway dimensions. However, Raytheon's project was a technical success in that the control software and in-vehicle steering mechanism worked. If they had kept the vehicle size & weight small, the guideway structure would have been cheaper, smaller and less obtrusive, and it would have been a cost success as well.
You're exactly right on the private ride and security features, although I'm not sure if ULTra has a panic button. There's an additional deterrent: Only the person buying the ticket knows the destination ahead of time. If you were a mugger or rapist, would you force your way onto a monitored Taxi2000 car without knowing where you were going, and the only control you have is to hit the Panic Button, which only means alerting the police? Of course you wouldn't.
You mount cameras in the station and inside each vehicle. Because a smartcard is used to program destinations, you can match the time of vandalism agains the log of smart card usage.
ULTra and Taxi2000 PRT cab drive and control mechanisms are inaccessible to the rider. If a car is unusable, it is sent to the maintenance barn, or if it breaks down it is pushed by another cab.
A PRT station is off-line on a siding, so a car stopped at Station X doesn't delay a car going to Station Y.
Boarding and unboarding doesn't cause congestion in a station because an empty cab is used by the next rider. If someone needs to arrive at a station full of empty cabs, the system knows and clears out a berth. The empty cab goes to a station where it is needed.
Serial vs Parallel: The main reason a subway picks up and drops off a crowd of people at the same time is because the schedule requires people to show up at an appointed time & ride together. PRT is on-demand, people ride when they want. Therefore, no crowds at stations, therefore huge stations aren't needed.
Try reading this paper:n d- Effect.pdf
http://www.taxi2000.com/pubspdf/Failure-Modes-a
pay attention to pp. 7-8 and 16-22
1.
The number of little driverless cabs is less than the number of cars because the little driverless cabs are reused by a succession of riders. Trips by little driverless cabs are shorter than trips by car, bus, or train because little driverless cab trips are nonstop.
Suppose each little driverless cab carries 5 different fares per hour. If a system has a fleet of 500 little driverless cabs, that's 2500 fares per hour. 2000 little driverless cabs = 10,000 per hour.
2.
The idea behind Personal Rapid Transit is to eliminate the need for transfers. There would be lots of tiny stations (smaller than a house) throughout the network, enough so that you would usually be within walking distance of your origin or destination.
3.
PRT systems are envisioned as INTRAcity or city-to-suburb PUBLIC transit. If you need to leave the area or travel 300 miles, then you'd take a car, plane or train. If you need to pick up lumber for a home improvement project, you'd use your SUV (if you have one) or have the lumber yard deliver. PRT is designed to alleviate urban traffic congestion by increasing transit's share of trips. PRT does not envision totally replacing cars.