Become a fan of Slashdot on Facebook


Forgot your password?
Math Australia Transportation Science

Aussie Government Competition To Predict Commute Times 79

databuff writes "Last week, Sydney's Minister of Roads, David Borger, launched a $10,000 competition to develop an algorithm that predicts commute times on a major Sydney freeway. The winning algorithm will be used to power predictions on the Sydney live traffic website. The hope is that the predictions will help commuters make informed decisions about when to travel and on what routes, lowering the intensity of peak hour traffic. In its first week, the competition attracted entries from more than 50 teams and 19 countries."
This discussion has been archived. No new comments can be posted.

Aussie Government Competition To Predict Commute Times

Comments Filter:
  • won't everyone just follow the same algorithm and end up with traffic peaking at some other time?
    • Well, it depends on the algorithm . . .

      double getTrafficLoad(Location location){
      return (new Random()).nextDouble();
    • by serps ( 517783 )
      Er, no.
      1. Not everyone has access to the web at work to find out. Often, the first you hear about traffic congestion is via the radio in your car.
      2. Not everyone has flexible work hours
      3. Not everyone has many choices of route home
      4. People are lazy.
      • I get the part about lazy.
        Apparently the minister of transportation and roads is too. Here in Denmark, we've had nice functionaing algorithms, based on the length of the queue, and the speed at which it moves. And with some nice LED signs that says "traveltime x minutes from here to whereever", and we've had those for.... Well, at least 10-15 years now. A tried and proven system, what's the point of asking for a new algorithm ? Does he want something that makes a GUESS rather than bases itself upon the fact

        • No he wants something that works with the higgledy piggledy layout of NSW roads. Maybe in Denmark your roads make more sense?

        • I think the government here is a little short on cash. The algorithm you use (and is probably wanted here) is possibly too expensive given the budget of the program.

          My guess is that they're trying to get it on the cheap ... $10K is cheap for a fully functioning algorithm and the full rights to use it.

          Congrats on having a fully functional government that is forward thinking.

      • I know I'll probably get shot-down but there's a very, very simple solution:

        Just as the internet companies have had to lay fatter "pipes" as consumer speeds have grown from 14k to 56k to 10,000k so too do the highway companies have to lay fatter roads to handle the growing population. Internet & roads need to be designed for the PEAK demand to handle all the "packets" full of data or people. But right now they are nowhere near that level - hence traffic jams and congestion.

        • by RMH101 ( 636144 )
          except if you build more get more congestion. Honestly. Demand expands at a faster rate. What's needed is a way of optimising what we *have* and encouraging car sharing/alternative forms of transport and ways of encouragign the above such as congestion charging, time-based road tolls etc.
          • >>>except if you build more get more congestion.

            Bullshit. Drive to the US midwest and you'll see a LOT of roads that were widened from 2 to 4 lanes wide as part of the Eisenhower Interstate upgrade..... and yet these roads are mostly empty stretches with little traffic. Therefore your conclusion that wider roads == more cars/congestion has been proven false.

            • by RMH101 ( 636144 )
              Google "induced traffic". There's a lot of transport research that goes on on academia, and its been shown that building more roads increases traffic exponentially. Obviously this applies to areas with high current concentrations of cars. If someone wants to build a 4 lane freeway in the middle of Hicksville for a kickback this may not apply, obviously...
            • by xaxa ( 988988 )

              In the UK, "reducing journey times" (or "reducing congestion", same thing) can't be used to justify building a new road. This is because the reduced journey time lasts a few months, by which time X more cars are using the road since people adjust their lives to use the available space.

              People are willing to spend N minutes travelling to work every day. If a journey would take longer they'll do something else (move house, move job, ask for different hours, etc).

              The same happens with public transport. About a

          • I was called in on a project to help reduce the problem of congestion in an Aussie city. We were immediately cut off at the knees ... apparently, congestion is seen, and used, as an indicator to determine the wealth of a nation (just as Internet penetration/uptake is). If you are a wealthy nation, you have congestion.

            The policy makers deliberately prevent you from coming up with an all encompassing solution.

        • Most NSW residents would agree with you in principle. However it's not very practical in many areas of Sydney - there's simply no space to put extra lanes on many roads without the Government reclaiming massive (and I mean massive) numbers of homes and businesses. Which would cost a stupendous amount of money and cause an absolute uproar.

          In fact for the last 10 years, most of Sydney's major road expansions have had to be underground for this very reason. See, for example, the M5 Tunnel or Lane Cove Tunnel (

          • I was talking about expanding roads OUTSIDE of the city, so that people (and freight trucks) can take these alternate routes (i.e. drive around the city) instead of being forced to go through it.

            There should be plenty of room in the surrounding suburb or rural areas to build new high-speed, limited-access roads.

            • Thing is I don't think there's much in the way of 'through traffic' that's causing these problems. Australia isn't like Europe or the US where there's towns and cities all over the place (meaning a lot of traffic in any location is simply driving through there to get to another place). 99% of the traffic in Sydney would be there because it is, in fact, trying to get to places within the city of Sydney.

              As it happens if you do genuinely intend to simply pass through Sydney (say, you're driving from Canberra t

              • Just to clarify, I'm not shooting you down and I agree 100% with your assertion that things such as roads should be overengineered. But Sydney's geography and demographics pose some issues that aren't simple to resolve, that's all. It can be done but it would be 10x as expensive as in a place that had substantial amounts of through traffic/was flat and had space available on the outskirts/didn't have everyone trying to get in nad out of a few small areas every day.

    • As I travel in Sydney traffic every day after 7 am till 9:30. just about every road into an out of the city does a great approximation of a Carpark so real easy to work out ravel times Afternoon peak starts at 4:30 and stays till about 6:30 pm I assume the difference is that the evening traffic Spreads out into secondary roads, where the moring in just funnelled all into the city
    • by Yvanhoe ( 564877 )
      That is why making such algorithm is not only doing prediction but also deciding on when you want traffic jams to occur...
    • Easy. I have a motorcycle. My commute time is extremely predictable.
    • Exactly, but that won't stop a planet full of petrol addicted monkeys from driving headlong to civilisation's peak oil fuelled destruction.

      The drivers are addicted to petrol. Traffic engineers are addicted to petrol. Governments are addicted to petrol votes and petrol excise. Of course, while the frog boils slowly, no effort will be made to escape from the water pot.
  • by MichaelSmith ( 789609 ) on Thursday December 02, 2010 @03:50AM (#34414630) Homepage Journal

    I used to do this for VicRoads in Melbourne. At peak times a queue builds up at the exit end of the freeway. Normally the city end during the morning peak. Travel time depends entirely on the length of the queue. Maybe RTA NSW should be paying VicRoads for the software I wrote in the 1990s. It can be a discount on the money VicRoads pays for SCATS.

    • Yes but if the traffic in Sydney is anything like the traffic on the Eastern Freeway (which I assume you're talking about) then it no longer behaves the way that your model assumes.

      During the morning peak (which now begins well before 7am), now that the EastLink tollway has been completed, there is a fair amount of traffic already on the Eastern Freeway before you even reach Springvale Road. As a result, a commute that used to pretty much exactly follow your predictions now is bumper-to-bumper from Springva

      • Most of my data from back then was for the south eastern arterial, which was where we had the high resolution speed and volume data. We weren't working with the Eastern much in those days. I take your point that the tollway has changed the game. It just means you have queuing at different points of the freeway. Generally where there isn't a queue traffic will flow at the limit. Where there is a queue: 30 to 40 km/h. Our real time data used to look like this:

        • 1500 90 10
        • 1000 95 12
        • 500 105 9
        • 0 101 8

        Thats Distance

  • Every life has a purpose. The purpose of yours may be to provide a cautionary tale for those who come after.

    The obvious correct answer to this question is an app that leverages sharing the GPS info of participants in order to deliver to the same participants reliable real-time info about traffic conditions. The fact that you're flying down this particular freeway at 100KPH right now is valuable information to people who also know of other routes to where they want to be where people are creeping along o

    • by jrumney ( 197329 )
      You mean something like this []?
    • by RMH101 ( 636144 )
      too complex. Just use a (GSM) cell phone's unique IMEI number and take the data from local cell towers to triangulate the approximate position and speed of all cell phone users on that network. Vodafone do this in the UK and it's pretty effecient.
      • You're both missing the point here. Knowing the current commute time is fairly easy. The tricky part is using the measurements now to predict the commute times an hour from now. My commute ranges from 20 up to 50 minutes or so. If I can look on a website and see in a half hour my commute will be a half hour shorter, I'll just wait. If it's still going to be long, I'd rather slog it out now and get it over with.

        • by jrumney ( 197329 )
          Data collection is an essential step on the way to accurate forecasting. Having real-time data also lets the model adapt - if historical data suggests that a traffic jam of this severity in this location at this time takes 3 hours to clear, then knowing that the traffic jam has formed is just as important as the historical traffic pattern information for giving you that accurate forecast.
  • I predict the traffic will be really busy between 7am and 9am, and then again from 4pm to 6pm. Where do I pick up my prize?
    • I predict the traffic will be really busy between 7am and 9am, and then again from 4pm to 6pm.
      Where do I pick up my prize?

      When you can predict how Sydney's failing infrastructure affects the commute. I was late to work the other day (on late shift) because a water main burst on Paramatta Road and they blocked the road citybound for a couple of hours while they sorted out the mess. Sydney's roads and transport (and other infrastructure obviously) are so poorly planned and maintained that any algorithm that was remotely successful would need several good random number generators.

  • the prediction will effect the prediction. they already know the peak hour, just encourage people to use other times for their commuting. like sleeping late!

    • by houghi ( 78078 )

      In Germany I have seen that people have much more flexible hours and to avoid being in traffic would start earlier. The result is that traffic starts earlier.

      • In Germany I have seen that people have much more flexible hours and to avoid being in traffic would start earlier. The result is that traffic starts earlier.

        Exactly, and so do traffic jams.

        As a rule, having modestly flexible hours (which is all that ever happens) you just add an hour or so to both the morning and evening rush hours.

        There aren't many jobs where your flexi-time allows you to work from three in the afternoon to eleven at night (or three in the morning to eleven in the morning) if your hours are normally 9-5

  • Who cares? I've seen lots of traffic sites. Even a GPS with an optional attachment for live traffic reports. You know what they always say? Traffic is pretty much where you expect it to be. Everyone knows where the traffic spots are, and sure enough that's where the traffic always is.

    You'd think the usual suspects would be totally against this sort of science. The attitude is "we should annoy the living shit out of drivers at every opportunity, this will certainly make them take the bus instead. Pl

    • During peak hours the train is the fastest and cheapest way to get from the burbs to the city in Melbourne.
    • by AmiMoJo ( 196126 )

      So rather than just predicting traffic they really need to find an algo to reduce it. In this day and age we should have networked traffic lights with computer control to constantly monitor and adapt to the situation in order to optimise traffic flow. Instead we just have dumb traffic lights that run on timers.

  • by PolygamousRanchKid ( 1290638 ) on Thursday December 02, 2010 @04:13AM (#34414700)

    A while back, some folks in France recommended that vacationers should drive after midnight to avoid the August vacation traffic crunch. The result? Tons of folks drove at 03:00 on toll roads . . . however, only one toll booth was open, as usual, for nights. So there were major traffic jams for hours.

    So I ask you, Professor John Nash, what should I do out of a Game Theory analysis? If everyone is told, to drive at a certain time, is it better for me, if drive at that time? Prisoners' Dilemma, on the roads.

    The economist John Maynard Keynes was once asked, if he thought that the stock market would go up or down. He answered: "It's not important what I think. I invest on what I think, what the most people think."

    • by houghi ( 78078 )

      And what with all people using GPS with traffic avoidance build in it. The basic thing that I see has happened is that the secondary roads are congested now as well. Overall traveling time has increased.

      Then there are people who work at home and do the school run for their kids, because their kids are lazy and don't want to sit in a bus. They do this at the busiest time and have to do the trip twice, adding to the traffic.

      • by Matt_R ( 23461 )

        And what with all people using GPS with traffic avoidance build in it.

        Does that actually work? I have a GPS with traffic, and live in Sydney. I'm yet to see the traffic avoidance actually work. It tells me there's traffic up ahead, when there isn't. It tells me there's no traffic when there's a crash up ahead.

        So yeah, it gets it backwards. The name of the traffic info provider is SUNA Traffic []. Guess what SUNA backwards is...

      • by drsmithy ( 35869 )

        Then there are people who work at home and do the school run for their kids, because their kids are lazy and don't want to sit in a bus.

        It's not that the kids are lazy, it's because Today Tonight and various similar trash have convinced them that if their precious snowflake is out of their sight for more than a few seconds between authorised locations, they'll be kidnapped, raped and murdered.

      • by tlhIngan ( 30335 )

        Then there are people who work at home and do the school run for their kids, because their kids are lazy and don't want to sit in a bus. They do this at the busiest time and have to do the trip twice, adding to the traffic.

        It gets better.

        A school sits at an intersection (with traffic lights and pedestrian-activated crosswalks).

        People constantly circle the carpark and make U-turns on the adjoining road because the pickup/dropoff area is full of people waiting. Meanwhile, across the street, there's plenty of

      • Then there are people who work at home and do the school run for their kids, because their kids are lazy and don't want to sit in a bus.

        As opposed to all the people driving to work who are lazy and don't want to sit in a bus?

    • So I ask you, Professor John Nash, what should I do out of a Game Theory analysis? If everyone is told, to drive at a certain time, is it better for me, if drive at that time? Prisoners' Dilemma, on the roads.

      Clearly, a random time from 00:00 through 23:59. Then traffic will be spread evenly.

  • How much does it cost to rent a bulldozer? "Never" could be an accurate answer, yes?

    - RG>

  • I've been doing that for years already on my iPhone.

    Maps - Options - Show Traffic

    PM me for my address to send the $10K

  • No offense, but if you want a good traffic prediction algorithm, please offer more than my shitty quarterly salary.
  • ... 7. See the fuckers responsible for the M25 computer system shot in their heads.
    • Yes, because obviously that's responsible for holdups, rather than the fact that there's an awful lot of people trying to travel through an awfully small amount of space.

      • by ghrom ( 883027 )
        If you have seen a 40mph speed limit at 3am when the road was virtually empty and imagined what's gonna happen there 4h later, you would be less sarcastic. If you have seen a 40mph speed limit and a warning about an accident, and after passing the place seen no sign of any accident whatsoever, you'd be less sarcastic. If you have seen the official statistics claiming 1.5% increase of throughput within the peak period and no mention whatsoever about the off peak period, you'd be less sarcastic. If you actual
  • I always wondered why automated twitter like systems are not used to solve this kind of problem.
    Here is what I would like: Have a device which measures the traffic at one point installed all across the highways (Simple IR device looking at the number of passing vehicles per second should do). These devices act as social beings, sharing their data to a server to give out to everyone. They should come cheap, especially while buying in bulk (Make them solar powered, if you want to). Let them transmit the dat
  • If ((Easter || Christmas) && road==F3)
    delay = 6 hours;
    (You know the traffic's shit when you can play cricket on the 3 lane each way freeway.)
    • by MrKaos ( 858439 )

      If ((Easter || Christmas) && road==F3)

      Ahhh I see, another suffering commuter from the Central Coast, I've refined your code.

      if ((Easter || Christmas) || (random(truckCrash > 1)) && road==F3) {
      delay = 6 hours;

      if (fireSeason && random(bushfire.anytime) {
      delay+=8 hours

      if (gawkingTouristNotPayingAttentionWhilstTowingACaravan > 1) {
      delay+=(gawkingTouristNotPayingAttentionWhilstTowingACaravan * 2hours)

      if (Friday.northbound.afternoon || Saturday.northbound.morning ||

    • (You know the traffic's shit when you can play cricket on the 3 lane each way freeway.)

      How come the trafic isn't moving if there's space for a cricket pitch?

  • Self defying (Score:5, Interesting)

    by Ignatius ( 6850 ) on Thursday December 02, 2010 @05:53AM (#34415066)

    Predicting commute times and keeping the results secrets vs. predicting commute time and putting them in real time on a public website are two completely different problems. The former ist simply about estimating an output parameter from a set of input parameters so it's basically about approximating a function. The latter contains a nasty feedback loop as the output paramter is in itself an input parameter as it influences the behaviour of the system, so you're basically looking for a fixed point where the publication of the forecast exactly repells as many drivers at it attracts - only these values allow for a stable prognosis. In economics this effect is known as Goodhart's law [].

    This means that the competition is about a completly different (and much simpler) problem to that which they are eventually trying to solve.


  • Maybe if the government wants traffic analysis performed for free (well, a small possible prize) they should be a little more permissive with street level GIS data. For those in the US the equivalent of the United States TIGER/Line data that anyone can download for free costs about $10K - 200K in Australia depending on usage last time I checked. It's handled by an unlisted public company called the PSMA that various state and other government departments have stakes in but they're every bit as greedy as a p

  • They should use the SimCity traffic commute system. In my city my current commute time fluctuates from 0.5 to 30.
  • by dltaylor ( 7510 ) on Thursday December 02, 2010 @07:30AM (#34415418)

    If the employers already allowed flex time, there would be a more even distribution of commuters over several hours, and the problem would smaller to non-existent.

    If the employers don't allow flex time (as apparently they don't, as least in useful numbers), it won't matter what information is available to commuters, they'll still have to be arriving at work during peak commuter density periods, and leaving at the corresponding end-of-work-day.

    Really want to fix the problem? Maybe you could have an auction every year for employers to buy their preferred work start hours. For 1000 employees, it might encourage them to save a few bucks by buying "off" hours.

    • Problem is, when everyone's working at different hours, communicating/coordinating with them is going to be nigh impossible.
      Especially in retail. The store might not be open during normal business hours, so you look it up... And it's only open when you're usually eating dinner/sleeping. Now what?
  • You don't need an algorithm for Sydney's traffic, just this

    echo car > /dev/null

    Can I have my $10,000 now please?

  • Their measurements will alter what they are attempting to measure. Announcing 'that route is gridlocked, use this route' merely results in two gridlocked routes.
  • Int(s0..sf,ds/v(s))

    Seriously, though, a traffic prediction algorithm that worked ok actually would be useful for designing changes to the roadway to reduce traffic, but I suspect that Australia is probably similar to the US: you probably know what causes the traffic, but for whatever reason, you're unwilling to do what it takes to cut it down. For instance, designing roads for peak capacity instead of average capacity, putting pressure on businesses to spread out work hours more to flatten the curve a bit,

  • A nitpick, but an important one: the government in question here is a State government, not the 'Australian' (Federal) Government. To be exact, the government of New South Wales (which contains Sydney).

    I'm sure most of you will roll your eyes but I can imagine Americans would have a similar reaction if an article called, say, the government of South Carolina or Idaho or Minnesota or something, 'the American government'.

  • The hope is that the predictions will help commuters make informed decisions about when to travel and on what routes

    Heh, I misread "commuters" as "computers" and immediately thought, "What? Australia's already rolling out computer driven cars?!"
    Then I read the rest, and I was disappointed. :(

  • I got so sick of sitting in traffic for hours that I moved to the city so I can walk to and from work, I've been doing that about 4 years, it's great - however I miss some things from living in the 'burbs - community, large garage, dog....
  • This is an Australian state government. What's worse it's the NSW state government. The problem is further exacerbated by it being TRANSPORT related. The end result will be that somehow $350 million will be spent before the whole thing will be abandoned as a bad idea.
  • Randomize both time and route. There are enough simple ways such that your route is only elongated by a small margin, but the peak congestion you see can be extremely reduced.

Genius is ten percent inspiration and fifty percent capital gains.