Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×
Communications Technology

Obtaining Real Time Transit Info? 41

Posted by Cliff from the when's-that-next-bus-due? dept.
remolacha asks: "I am working on an EU initiative to create small-scale demand-driven public transit (minibuses + GIS + logic + SMS gateway for people to request transit) in rural Ireland. Our plan is to connect several small towns that have no public transit to a bus stop served by a national carrier, so as to enable residents of these towns to reach the two neighboring cities (20 miles in each direction, which have jobs, universities, etc). The thing is, no one wants to be dropped at a bus stop without knowing when the next bus is coming (buses don't come that often here, and schedules aren't reliable.) We'd like to provide real-time information on when the next bus is coming, as you see in some large cities. the problem is our project is so small, none of the companies that make these systems will even give us a quote. so my question for Slashdot is this: is there a homebrew or inexpensive way to set up a display at a bus stop that will tell users when the next bus is coming from a city 20 miles away? There are hills involved, so line of sight isn't an option."
This discussion has been archived. No new comments can be posted.

Obtaining Real Time Transit Info? More

Obtaining Real Time Transit Info?

Comments Filter:

  • GPS and a mobile phone (Score:5, Informative)

    by gkelman ( 665809 ) on Wednesday June 16, 2004 @06:15AM (#9440084) Homepage Journal
    GPS unit from RS plus some bits to tie into a GSM phone and send GPS co-ordinates over GPRS every few minutes. The bus stop's sign would just download the information from the same server you uploaded the GPS co-ordinates to.

    Simple and used by numerous people / companies already.

  • Ireland Bus!! (Score:3, Interesting)

    by justinmc ( 710870 ) on Wednesday June 16, 2004 @06:29AM (#9440130)
    I grew up in Schull, West Cork - don't talk to me about the Bus schedule from there.
    Two or three buses a day out of the place.
    One on a Sunday! The day I got a car was one of the happiest in my life - the sad thing is I prefer to use public transport, but it just isn't an option in some parts of Ireland!

    Jay

  • Are you sure Real Time is the way to go? (Score:3, Informative)

    by trevry ( 225903 ) on Wednesday June 16, 2004 @07:23AM (#9440273) Homepage
    Are you sure that real time data is the way to go with this project?

    The arrival time would have to be calculated from the speed of the bus, either averaged or real time. Both would be inaccurate in the context of Irish roads. Irish road distances, both on maps and signposts, are INCREDIBLY wrong. Not inaccurate, just plain wrong.

    I've driven all over Ireland and the best way to know how far and how long it's going to take to get from point A to point B is to use experience.

    Just get the drivers to punch in an estimated arrival time. It'll be right to within five minutes. Their tachys will show whether they were right or wrong.

    The drivers of the buses are going to have driven that road before, in those weather conditions, at that time of the day, in the current traffic conditions. They'll know if Mackey the dairy farmer takes his cows across the Miltown Malbay road at 11.30am EVERY day. They'll know if the Oughterard road floods at this time EVERY year and you have to go "the back way".

    If you're standing at a stop there is NOTHING more annoying than seeing a delay tick up. If you use realtime data then the times are going to vary wildly depending on the speed of the bus at various moments. If you average it will just be plain wrong, in the west of Ireland there are a lot of sections of bendy roads followed by straight road, this is because a lot of the roads were laid out during famine times. The roads just meandered along, not really going anywhere, the workers didn't care how long they were, they just wanted to get paid and fed. There are roads that go on for miles before just stopping dead, in the middle of nowhere.

    For my mind, the best solution would be to let the drivers estimate. It would also be a better solution cost wise. Just get the estimate transmitted to the next stop.
    • I was reading this and thinking that asking the drivers made sense, but when I read, "Just get the drivers to punch in an estimated arrival time. It'll be right to within five minutes. Their tachys will show whether they were right or wrong," I started thinking about error correction. The past is the best place for real data. You can start with estimates, but you need to store the data, so that as the system ages, it can learn from experience, and get better at predicting the time based on the distance. Sur
    • I've driven all over Ireland and the best way to know how far and how long it's going to take to get from point A to point B is to use experience.

      And why cannot software do this? Don't use time/distance/speed. Use experience: the average of the last 100 journeys from to here was Y minutes, so when bus reaches , advertise it in Y minutes time. Don't average over too large a window, so that it automatically adjust itself when busses travel slower in winter or there are roadworks nearby. Requires modest amo
    • besides than that.. It really isn't THAT hard to get the busses arrive on schedule(or 1-2 mins late at maximum, it is important to not leave the stop early though.). especially this estimation is easier on rural areas.

      (hell, the busses arrive on time more often than trains.. and this is in rural Finland even during winter)

      besides, if there's just 1 or 2 bus per day the thing that you need to know is something akin to "is it 12 o clock yet?".
      • In Ireland it is impossible to get Buses to run on time. They just don't. They never have and never will.

        At least by making the drivers someway responsible (by making them notify the stops) the public can see when the damn things WILL eventually arrive.

        I travelled on the Irish bus and rail network for over twenty years until I got a car. It's the worst in the world. I've been on buses that were two hours late leaving or three hours late arriving. I've waited for buses that simply never came. I've even wai

        • A note about my Greek Bus experience that might apply here:

          After experiencing the Greek (specifically, on Crete) bus system, I made up a new phrase and use it occassionally to describe the actions of my 2-year-old toddler: Having The Mind Of A Greek Bus Driver.

          This translates to, "Not knowing what you're going to do until AFTER you've done it".

          The bus driver sometimes stops, sometimes he doesn't, and you have to wait until after he was supposed to stop to find out what the decision was, even if you ask h
    • The arrival time would have to be calculated from the speed of the bus, either averaged or real time. Both would be inaccurate in the context of Irish roads. Irish road distances, both on maps and signposts, are INCREDIBLY wrong. Not inaccurate, just plain wrong.

      If the system was designed to be intelligent and learn from previous journeys then it would become increacingly accurate as time progresses. It would be good if some form of traffic info was also used however I doubt that this is collected.

      If
    • I agree that allowing the driver to punch in estimates is a great idea. However, why is the following the only way of doing things?

      The arrival time would have to be calculated from the speed of the bus, either averaged or real time.

      If you collect data for a while before you put up the sign, then you should be able to establish reasonable milestones. E.g., when the bus hits point A, it's within 20 minutes 95% of the time and within 15 minutes 80% of the time. This will of course vary according to outside

  • Really need more info but, here goes: (Score:4, Interesting)

    by stoborrobots ( 577882 ) on Wednesday June 16, 2004 @07:24AM (#9440278)
    Ideally, you would say what kind of connectivity is available in the area you are working on. For example
    • Is there cell-phone coverage throughout the area? Is it GSM or CDMA? (If it's GSM, you could acquire receive-only SIMs from a provider, and send updates from the bus via SMS).
    • Is there cabling in the area (telephone/electricity) which is easy to tap into? (Could you make a deal with the telco to provide a line to the bus stop?)
    • Are bus stops located near other facilities which could jointly use services (eg pubs/service stations which could share an internet link with the bus stop)?
    • Are there any locations from which a radio antenna could transmit to the entire route (or a significant portion thereof)? (Could you set up a transceiver on a hill, receivers in the bus stops, and a transmitter on the bus?)
    • Is the area geographically small enough? How many roads are involved? How far do the roads deviate from one another? (Could you run transmitters along one road which could be received from a large number of other roads?)

    Just a few thoughts off the top of my head. HTH. YMMV. HAND.

    • If there is cell phone coverage, there are several companies that sell GPS devices that used a celluar modem to periodically transmit location data, usually to a central server (read: subscription fee) where you can either access dynamic maps with vehicle location or download streams of location info (using webservices or XML).

      In the U.S., I know of at least one company (Cloudberry) that uses a dual celluar/satellite technology (use celluar if there is a connection, if not, use a satellite uplink).

      These s

  • Wi-Fi? (Score:2, Interesting)

    by peterdaly ( 123554 )
    My not knowing the area is an understatement, but here's an idea.

    Wi-Fi at (or around) each stop. Have a device, old laptop etc on the bus that sends out a "here I am" type ping through Wi-Fi and the internet which is received by a central server which can identify the access point of origin (by IP address maybe?). As the bus drives around, a "location" will be sent at each stop to the central server, which then can be used to calculate and transmit an ETA to the other stops.

    As a bonus, you may be able t

    • Re:Wi-Fi? (Score:1, Insightful)

      by Anonymous Coward

      Hello?!?

      The last line of the article said:

      There are hills involved, so line of sight isn't an option.
      Good luck in getting quality data.

      If all you need are simple pings, try something like HAM frequencies. You don't need the throughput, and line-of-sight is not an issue.

      • Original post: ...not knowing the area...Wi-Fi at (or around) each stop...central server which can identify the access point of origin (by IP address...which then can be used to calculate and transmit an ETA to the other stops...other "hotspots" along the route if you need additional tracking points...

        Flame-brained reply: Hello?!? The last line of the article said: There are hills involved, so line of sight isn't an option.
        Good luck in getting quality data.

        My Insightful response: Hello?!? Every oth
  • If each bus had a CB radio, they could periodically announce their position to a central office. Then people can call the central office to check where the bus is and get ETAs. Maybe local radio stations could announce bus information along with traffic/weather/news.

    Sure, it isn't tecnologically cool, but most public transportation already has an office, a phone number, and maybe a radio.
    • A number of the companies involved are sole-traders effectively. There is no 'central office' at this point. Perhaps you could organise the local community groups to provide something along these lines? Not unreasonable given that public transport is pretty fundamental to an areas vitality.

  • A suggestion: Have bus drivers use cell phones. (Score:3, Interesting)

    by Futurepower(R) ( 558542 ) on Wednesday June 16, 2004 @08:13AM (#9440462) Homepage

    A suggestion:

    Maybe the bus drivers all have cell phones, already. They could program a speed dial button for a voicemail answering system. When they arrive at a stop, or a landmark, they could press a speed dial button and say, Bus 6 arriving at stop 17, [town name], [time].

    If bus drivers don't have cell phones that work in the area, they would have to call from pay phones along the route. This would be convenient if they were stopped for a few minutes. It is possible to buy inexpensive devices which play back a series of tones, so that the drivers would not have to enter the digits by hand.

    Bus riders would call the voicemail number, and hear the recorded messages. The speed dial buttons would not only call the voicemail number, but also pick a voicemail box, and erase the previous message. So, bus riders would choose the voicemail box corresponding to their area.

    The disadvantage of this is that some people might have to pay a toll charge to call the voicemail system. The advantage is that the service might be close to self-financing.

    Some answering service systems have almost unlimited mailboxes, so other messages of interest to the community could be programmed, also. For example, a sign at the town hall could say, "If you want to know the location and date of the next [town name] town meeting, choose voicemail box 230." Or, a sign could say "Anyone wanting to know how Mrs. O'Leary of [town name] is doing in the hospital, call voicemail box 1472, and enter access code 241."

    People say good things about the free Asterisk [asterisk.org] telephone system software. Here's a quote from the web site: "Asterisk provides Voicemail services with Directory..." You would need only an old computer, since Linux and voice don't require much CPU speed.

    One computer could handle a very wide area, and many bus routes, I think, because each message is very short.

    Messages could be more detailed than just time and place. The bus driver could say, "Stopping for unscheduled maintenance at [town name]. Running late about one hour."

    With considerable programming, people could leave their number to be called when a bus arrives at a particular town just previous to theirs.

  • A source for info (Score:3, Informative)

    by UncleRoger ( 9456 ) on Wednesday June 16, 2004 @09:42AM (#9441198) Homepage

    You might talk to the folks at NextBus [nextbus.com] who offer that sort of info for San Francisco's Municipal Railway [sfmuni.com]. (You can check my stop [nextbus.com] if you want.) They are probably too expensive for your situation (I suspect they're out to make money) but they might at least have some advice for you.

  • I don't know of a homebrew setup, but I imagine there's a lot of work involved.

    but to see what a commercial solution is like, check out the "web watch" on MY busline. [duluthtransit.com] :)
  • the more your solution sounds like a taxi service.

    I think you are trying to be too high tech in your solution. The way I see it, based on your post, there are two options.

    1. If it is a minibus that runs a scheduled service, then put up a poster with the schedule on it.

    2. If there isn't a fixed schedule but, rather an on demand arrangement, then use a central dispatcher like taxi services do. The rider calls the dispatcher to request service and the dispatcher tells them when the bus will be there.

    Electr
  • Take a look at their MyBus [mybus.org] website, where you'll find this [mybus.org] link to their "Variable Message Sign" project. There's a fairly complicated software stack behind this, and I have no idea what they'd charge you for it. Also, a real-time data feed from the transit fleet is required, which is a huge, expensive requirement. Good luck!

    Here's [PDF file] [gpsworld.com] a good article about the technology.

  • I work for a medium size public transit company in Southern California. All of our busses are fitted with a gps and Siemens TransitMaster system.

    We are starting to implement signs like you are talking about at some of our bigger stops (ie, "Route 45 - Next bus arriving in 10 minutes") They also have a web module that you can use on your company's web site.

    I'm not sure how much the system costs, but I know that they supply some of the other municipal bus systems around here that are much smaller than u
  • A colleague and I recently did something similar for a company in the US who wanted to track shipments, but it should work similarly for you as well. Here's how it worked:

    Each pallet was tagged with an 802.11 RFID-esque tag which would broadcast it's info every 30 seconds or so for it's lifetime (about a week). There was a unit placed inside each truck which could capture the RFID info and spool it up, then use a GPS to grab it's position and connect out via the GPRS network to our server with all its i

  • .. I'm not sure I'd use them they are a really small company.

    http://www.nextbus.com/

    They company is called nextbus and they are being used a little in the bay area. Basically they use gps to track the busses and they you are supposed to be able to see when then next bus is going to arrive by looking at their map. They have wireless access to the system. I think this would work for you, but I also believe that this is a small company that has no european ties, so they may not be able to do everything.

  • But for real-time trafic info rather than for transit. Tripcheck [tripcheck.com] recently even added a real-time map of the only real metro area, Portland, with estimated speeds on all of the interstates.
  • Lookup APRS from the HAM radio world. GPS plus radio equals real time position reporting. Some of the equiptment will also act as data repeaters so infrastructure is low and you may be able to use the HAM frequencies. Equiptment is becoming off the shelf. Add a laptop and a screen at the bus stops and you get a realtime map of the area with the location of the busses.
  • ...and the link makes me doubt it, you could email me at kevin@doolin.com. almost all of our work involves linux, financial transaction systems, sms and web front-ends.

    as i live in north co. galway, i'd be eager to see you succeed.
  • Let the users ask when the next bus arrives by SMS. Already in use in Leicestershire [engadget.com].

Slashdot Top Deals

Software production is assumed to be a line function, but it is run like a staff function. -- Paul Licker

Close