A new idea for public transportation

December 11, 2003 8:47 PM

Intelligent Grouping Design is... a new idea in public transportation. With many vans out and about town, a passenger can be quickly picked up wherever he happens to be and just as quickly conveyed to his desired destination. Via the cell-phone, people call into the central computer with their current location as well as their destination. The computer finds the nearest van whose route is also the most closest to the passenger's destination. The computer then modifies the route slightly to accommodate the new passenger's pickup and dropoff locations. The drivers don't have to exert themselves mentally on figuring out each route change as the vans equipped with satellite guidance technology.
posted by gregb1007 (28 comments total)

For years, I've thought this is the way the bus systems should be run.

In my town -- 30000 people, it takes about 10 minutes to drive from end-to-end, and I regularly walk across town in all of forty-five minutes -- we have these enormous damn buses that are generally dead freakin' empty.

But there they go, 'round and 'round, on inconvenient routes at awkward times, making themselves useless.

So if we've got so few riders, we might as well have short buses on-call. And if we can do it with computer mapping, so much the better. Lord knows I'd take the bus *A LOT* if it were that convenient.

I suspect many, many others would also make use of it.
posted by five fresh fish at 9:29 PM on December 11, 2003

...a new idea in public transportation

Or a really expensive taxi cab.

Is there some reason why we need satellites to get a ride home? Seems like killing a mosquito with a bazooka to me.
posted by spilon at 10:11 PM on December 11, 2003

Anything that reduces private car use has to be a good thing. I think it's a great idea, with analogies to car-sharing and minibus "systems" in place in many cities and rural areas in Asia and elsewhere. Could it work in London? Probably not the best place for a trial, given the size of the place and the inadequate road system, but if it could be made to work here it could probably work in any major city, if the right investment were made.
posted by cbrody at 10:16 PM on December 11, 2003

I agree with spilon. It'd be a lot more efficient to use GPS location technology or even cellular triangulation than satellites.
posted by LukeyBoy at 11:07 PM on December 11, 2003

We've had this in the States for years. It's called disabled paratransit, and it is a miserable failure at delivering people in a reasonable amount of time at a low cost.

The program is called Trapeze PASS and is used by many transit agencies to do the disabled and senior dial-a-ride which many cities have. The average cost for these systems in Los Angeles is about $27, or $2.70 a mile. This is a shared ride system and it costs more than a taxicab, which is $2 a mile.

That is why that these sorts of systems are often tried in suburban areas but ultimately discontinued, at least for the general public, because they cost so damn much. Even if the bus is empty, at least you know it will be there at some time and will get to a destination in so many minutes. Such a demand based system interjects so many variables that the service becomes unreliable.
posted by calwatch at 11:16 PM on December 11, 2003

"I agree with spilon. It'd be a lot more efficient to use GPS location technology or even cellular triangulation than satellites."

you should email them and tell them, i'm sure they've never heard of any of those, and just arbitrarily chose satellites.
posted by luckyclone at 12:23 AM on December 12, 2003

Well, I think so far what has made these systems too expensive is the fact that the people would receive requests at call centers and call the drivers manually to tell where to pick up various people. That means you have to pay dispatchers to do the communication and that passes the cost to the rider. Once you have a computer doing this automatically, you save money on dispatchers, and the rides are much cheaper.
posted by gregb1007 at 12:25 AM on December 12, 2003

London's Dial-a-Ride recently celebrated its 21st birthday, and has 315 vehicles which make 1.3 million passenger journeys per year. It is struggling to keep up with demand. It's subsidised of course, as public transport has to be for it to work (in the vast majority of cases anyay.) The total cost appears to be around £12 per trip (with the passengers contributing 60p per mile or less), which is about the equivalent of a 3-mile taxi fare. So maybe it's not a lot cheaper, but it's certainly more environmentally friendly. And they're using humans, not software, to make the scheduling decisions.

So the economies of scale could probably make this work here for the general public, with enough public funding to get it off the ground. A lot of people would probably happily pay the equivalent of a taxi fare if it got them to work & home on time, above ground and in comfort, while making them feel good about themselves for taking public transport.

LukeyBoy, spilon, how do you imagine that GPS works? On preview, luckyclone beat me to it.
posted by cbrody at 12:27 AM on December 12, 2003

I had this idea more than 10 years ago, but at the time GPS wasn't quite ready for public consumption. I knew I could do it with computers and automation though. (It's not a far stretch, I'm certainly no evil supergenius.)

Even on normally routed buses, we need more information technology. They already have GPS on most Muni buses, they need to pass that information to the commuter. (It's easier on trains, obviously.) We need cheap, durable, solar-powered and wireless infokiosks at the stops. 2-6 lines of LCD would do just fine. (Web browsing if only to a realtime schedule/map intranet would be heaven.) Just tell me if the damn bus I was transferring to has passed or not yet. We need this information to route to displays *on* the buses. ("Am I going to make my transfer to route A? Should I stay on this bus and transfer to route A1 or go onward to route B?")

We *need* just-in-time smart systems for public transport if it's ever going to compete with private transport and be at all attractive to your average consumer. Most drivers that refuse to take public transport seem to be either too moronic or too damn lazy to even be bothered otherwise. "What do you meeeean I have to waaaalk a whole half a bloooock to the stop? What is this transfer nonsense? I caaaan't read this scheduuuule! Oh, to hell with it. FTW. I'm driving. I don't care if it bankrupts me or kills me or anyone. Mememe!"

Make it happen, transport people!

Sign me,

Pedestrian/cyclist/straphanger for life.

(You pity me now - needlessly - while I laugh at you stuck in traffic sucking down CO, dying of stress, wasting your money on an illusion of speed, and getting fat and lazy. Don't pity me. I'm going to outlive you, smoking or not. And I'll sure as hell read a lot more books then you, and meet more interesting people than you, and see more sights of my fair city than you, sitting there on that lame little bus.)
posted by loquacious at 4:13 AM on December 12, 2003 [1 favorite]

"....That is why that these sorts of systems are often tried in suburban areas but ultimately discontinued, at least for the general public, because they cost so damn much."

Well, the cost of dispatchers might be a factor - "I think so far what has made these systems too expensive is the fact that the people would receive requests at call centers and call the drivers manually to tell where to pick up various people." (gregb)

But cbrody hit the nail on the head - "the economies of scale could probably make this work here for the general public, with enough public funding to get it off the ground." : until user density grows high enough, such systems will be inherently inefficient. But - once user density crosses a certain threshold - these system will be more efficient than our current passenger owned/driven vehicle system.

This post is really about Linear Programming, and there are already a number of commercial software packages - and even some open source linear programming software - designed for various types of optimization, including route optimization.
posted by troutfishing at 6:34 AM on December 12, 2003

This sounds like it would have a very narrow range of usefulness. Too few riders, and you'd be just as well off using taxis. Too many riders, and the system falls under the weight of its own complexity; you'd be better off with a predictable schedule and route.
posted by ook at 6:55 AM on December 12, 2003

"...you have to pay dispatchers to do the communication and that passes the cost to the rider. Once you have a computer doing this automatically, you save money on dispatchers, and the rides are much cheaper."

If it got to that point, don't you think it would be pretty surreal to be the driver? You'd never really know where you're driving. As people called in to get picked up the route would be constantly changing and the driver would just be blindly following directions.
posted by fletchmuy at 7:05 AM on December 12, 2003

...all details of all taxibus journeys are logged on the controlling computer system. These details include the identities of the driver and all fellow passengers, the exact route traversed by the taxibus, the place and time each passenger boarded and alighted, and for extra safety, video surveillance inside the taxibus.

Okay then. That they don't care about customer privacy is not surprising, but it's a bit odd that they go out of their way to proudly state that you won't have any.
posted by sfenders at 7:13 AM on December 12, 2003

Sounds like SuperShuttle, with variable endpoints. I've always thought that service was incredibly inefficient even with fixed endpoints and substantial time between call and pickup. It's hard for me to imagine that making the equation more complex would (by itself) improve the service above SuperShuttle's.

For that matter, getting a cab in San Francisco, which has the same use model and a much simpler decision (nearest empty car is assigned to the fare), doesn't work at all. Again, simpler problem, poor results.

I'd be more into the idea of replacing large busses with many more small vans, and having their routes be set but more varied than the large busses' are. I bet you could make a more efficient public transportation service that way than either a large-bus system or the just-in-time route scheduling system described in the post. (It would cost a lot more than large busses, though, since labor costs would be double or more.)
posted by precipice at 7:55 AM on December 12, 2003

"Too many riders, and the system falls under the weight of its own complexity" - ook, I don't know if you checked out my linear programming links, but this is actually a standard sort of business problem - long taught in business schools - called the "many drivers/many cities problem" (or comparable names). Large scale businesses with large fleets and complex routing demands tend to turn to linear programming to optimize their routing - linear programming takes a lot of processing power, sure, but it's been proven to work in real world situations.

sfenders - I accept the privacy infringement which my "Speedpass" - for paying tolls electronically via a little transponder clipped to my car's windshield - implies. I'm not totally thrilled about it, but I like the time savings of not waiting in lines at tolls. That paragraph you quote is a bit over the edge though, I agree.
posted by troutfishing at 7:58 AM on December 12, 2003

When I first read this post, I immediately thought of Badger Cab in Madison, WI. When I'd visit friends there in college years, I was amazed at that service. Everyone in town knew the phone number, and it seemed that no matter where you called from, the dispatcher would tell you to be outside in two minutes or miss the ride. A Badger Cab would roar up within moments, often with 2 or 3 other people aboard. Somehow, their destination was always on the way to yours. IIRC, they didn't run a meter, just charged two bucks a head. The drivers were incredibly motivated and efficient. It was exactly like taking a small, fast bus with on-the-fly routing.

They had a sharply defined service area, as it would clearly get all screwed up if they tried to deliver people too far, but you could easily and cheaply get anywhere a student would think of going.
posted by Tubes at 9:39 AM on December 12, 2003

I immediately thought of Badger Cab...

Sounds like a great system to me. Do any other smaller cities have this kind of shared-ride cab system?
posted by jacobsee at 9:54 AM on December 12, 2003

Too many riders, and the system falls under the weight of its own complexity; you'd be better off with a predictable schedule and route.

Well, no, that's sort of the point. As the number of riders and buses increases, the system becomes both more efficient and more predictable.

I think it's a pretty great plan - it seems like the idea has been floating around for a while in various permutations, but we're just hitting the point where the technology to make it work properly has become economically feasible.

None of the superficially similar, unsuccessful systems that people have been pointing out in this thread have all of the elements that are necessary to create the network effects that will allow this system to scale gracefully.
posted by jeffj at 10:53 AM on December 12, 2003

Most drivers that refuse to take public transport seem to be either too moronic or too damn lazy to even be bothered otherwise.

For me, I avoid public transport because in the lansing area, it can take less time for my wife to drive from detroit and pick me up than it can take for me to catch the bus, complete the 1 transfer necessary, and go to a destination that's litteraly 5 miles from my house. (I kid you not, this is exactly what I experienced the last time I took the bus, and it's not like the bus even had to go out of the way, the route you would drive is the same as the route the two buses take. The bus I was transfering to pulled up about 2 minutes before my wife arrived.) Home to work, about a 15 minute drive, would take between an hour and an hour and a half to do on the bus.
posted by piper28 at 12:51 PM on December 12, 2003

If it's only 15m driving, why not ride a bike? Wouldn't take you more than about 20m. Power-walking would take less than an hour.
posted by five fresh fish at 1:53 PM on December 12, 2003

15 minutes driving == 20 minutes by bicycle? I'm guessing that either your car is stuck permanently in second gear, or your name is Lance Armstrong.
posted by sfenders at 3:40 PM on December 12, 2003

I can get to many places faster by bike than anyone I know could by car. (15 miles = about 30 minutes on flat land by bicycle at an easy pace.) This is often due to traffic conditions, but Los Angeles is famous for its traffic. (And geographic sprawl, which certainly makes being a pedestrian and straphanger here difficult, but more on that in a moment.) Another factor to consider is that a bike can take shortcuts a car cannot.

Another consideration is that I don't have to spend time driving in circles to look for parking in high traffic areas, which saves considerable time, not to mention money.

Having a family with children makes choosing public inconvenient at best, but not impossible. And it is a sacrifice one chooses to make, but in making it you can be healthier, happier, and less stressed out when done smartly and with planning.

A major factor in all of this is geographic sprawl and suburbia itself. See James Kunstler's The Geography of Nowhere for one take on its ills, and Michael Harrington's The Other America for a take on how poverty, suburbia, and inner city decay factors into this and other problems. Fleeing the city cores instead of improving them damages us all, wastes valuable resources, wastes arable land and fosters an insular, isolated car-centric society.

One of the amazing things about community transport is that it helps build real communities and communal awareness rather than simply providing a stark facsimile of one.

Again, committing to taking public transport is most often a sacrifice, but the only thing you should really be sacrificing is convenience. Granted, that difference between 15 minutes and 90 minutes is a large sacrifice, but I'm going to have to call a duck a duck and call that a form of laziness. (I realize time with your family is valuable, but I honestly hope you're spending time with your family and not simply rushing home to watch TV. I assume both sides of the ideal.) That, and it sounds like your public transport system is teh suck, but you have the power to change that. Join a bus riders union, lobby your city, get involved and get active.

Think about it for a moment. Think about all the resources tied up in public roads. Think about all the materials and energy in private cars. Think about the wasted energy of hundreds of millions of cars idling at stoplights.

Then think of the fantastically efficient and convenient public transport system that could be built with a fraction of those resources. As a collective society, we are entirely failing to self-organize and do things in any sort of an efficient or convenient manner.

Now think about the history of the automobile industry and the influence it has exerted over the last century to actively limit our public transportation and alternative energy options. (The Geography of Nowhere gets into this, and Google is your friend.) This isn't paranoid time-cube tinfoil hat stuff about cars that run on water, this is confirmed histories of companies like Goodyear buying New York City's perfectly functional electric trolley system and running it into the ground to sell more tires. We're getting screwed in the pooter, people.)

Think. Visualize an ideal future. Act.
posted by loquacious at 4:18 PM on December 12, 2003 [1 favorite]

loquacious, i give you a high five. no - a high ten.
posted by luckyclone at 6:45 PM on December 12, 2003

75 minutes of my time is far more valuable to me than any consideration of public transportation. I'm also not real convinced that a more efficient method of public transport between me and work would really be likely to ever exist (and quite frankly, I'll be damned if I'm going to invest the effort to try to improve it, since I'm not really a fan of public transportation anyways). The reason I'm not convinced it would get much better is because the way the lansing area is laid out, most of the routes run radially into the city, and it kinda makes sense that way. The problem comes when you have to go from one part near the edge, to another that's not near the city center, the times get real bad. Meanwhile, on my car I go through largely rural areas, and it just wouldn't make much sense to run a bus route through that direction. For the record, it's 12 miles that I have to cover by car (the bus route might as much as double that).

Biking isn't really an option in my opinion, largely because I don't want to sit through work all sweaty (and I'm sure the people I work with would agree with me). The roads I'd have to travel aren't all that great for biking either, fairly narrow, 55 mph roads.
posted by piper28 at 6:58 PM on December 12, 2003

Many progressive companies these days provide showers for cyclist commuters for just that reason. It's an easily solvable problem, and not one that should be considered an impossibility at all. I know cyclists that get gym memberships to gyms nearby their work. They bike there, shower, change into their monkey-suits, and work the day away.

I once worked in a very high-end corporate environment that was stunned to learn that I biked/bussed to work every single day. Stunned to the point of almost laying me off because they somehow thought I was less reliable or some other ridiculous notion. Three days later we had a series of statements of qualifications that were due immediately, and we couldn't finish them in house. We needed them done in minutes, not hours. I grabbed the source documents, hit my bike, and was at Kinko's in under five minutes. I returned in less than thirty minutes with the completed documents much to their disbelief.

Road safety is also a valid argument, but consider the source, and consider what can be done to change it. In San Francisco there are cyclist activist groups that lobby for restriping of lanes for cyclist use (with very stiff penalties for infringement of right-of-way by automobiles), even going so far as to publish easily readable topographical maps to help cyclists route themselves efficiently around those damnable hills.

The reason you should ultimately invest time and energy into public and alternative transportation is for your children and the children of your children. To ignore this is thinking selfishly in the only short term, thinking of only your own comfort and convenience, thinking of only the smallest of pictures.

Hydrocarbons are also quite valuable for things outside of fuel - things we haven't even discovered yet and things we have - and when they're gone, they're gone. We shouldn't be burning this stuff heedlessly, we should be using it for the plethora of materials it is capable of producing. (Responsible use of that as well is a whole different rant. :) )

Also consider the possibility the reason we're in Iraq and the middle-east at all - and the reason we haven't attacked Saudi Arabia, the real source of the 9-11 terrorists, like we have Iraq and Afghanistan - is perhaps indeed oil. We need to run Halliburton and its no-bid contract up a flagpole.

As well, consider the mortality and injury rates of private transportation, drunk driving not included. Automobile incidents kill more people than illicit drug use, but we don't see people getting their panties in a bunch over that all the time, do we? (And driving could arguably considered a drug that releases endorphins and is even physically and psychologically addictive, above and beyond the metaphoric.)

I would not be so opposed to private transportation if we were able to do it sustainably, wisely, and cleanly. Transport should not be a fashion statement with built in failure rates.

The fact of the matter is that the majority of people living in developed urban and suburban environments don't need to use their cars every single day. You don't need 3,000 pounds of steel and plastic to go to Blockbuster to pick up a video and box of microwave popcorn.

Take a walk. Ride a bike. Skateboard. Rollerskate. Heck, do a silly walk backwards.

It's pleasant, it's good for you, good for me, good for all of us, and you'll be surprised at all the details you miss when zooming by at 50 mph.
posted by loquacious at 7:42 PM on December 12, 2003 [1 favorite]

15m driving in the city = 20m by bike, yes. In your car you certainly are not averaging more than 30kmh in city traffic. On a bike you can easily maintain 20kmh, plus you get to cheat by lanesplitting any lineups at stop lights.

I believe, though, that in most downtown/urban areas, you'll find that riding bicycle is frequently faster than driving. There is a reason bike couriers are popular.
posted by five fresh fish at 9:42 PM on December 12, 2003

...I guess it depends on the city. I do tend to forget how bad rush hour is, since I'm almost never in it.

Anyway, IGD doesn't sound too intelligent to me. We already have tens of thousands of cars on the road, many of them going mostly empty. My preferred imaginary future intra-city transportation system is an electronically-asssited version of hitchhiking. Same deal as IGT basically, except use the drivers we already have, and keep the system open, free-market style.
posted by sfenders at 6:07 AM on December 13, 2003

Too many riders, and the system falls under the weight of its own complexity; you'd be better off with a predictable schedule and route.

I worded that poorly; I was thinking of complexity from the riders' perspective, not the system as a whole.

Think of this from the passenger's point of view, assuming the system is getting heavy use: you have to call in every time you need a ride; you never know exactly when a ride will be available; and you have no way of predicting how long it will take, because the routes change constantly. I think the average commuter -- assuming he's disposed to use public transport in the first place -- would rather walk half a block to the bus stop, where he knows the bus will arrive at 10:04 and take 32 minutes to get him to work, than to call in for a ride every morning and hope it all works out.

If Rider A gets stuck on a long, zig-zag path that saves a few minutes each for riders B, C, D, and E, then rider A is going to be late for work, and is going to take a car next time. On average everybody came out ahead, but you've just lost a customer and generated bad word-of-mouth.

Yes, on average, assuming enough number crunching, the rides are going to be shorter and more efficient than a traditional public transportation scheme would be -- but for any individual rider, I imagine the lack of predictability would be a huge turnoff.
posted by ook at 3:29 PM on December 13, 2003