The transit-obsessed at the UW are considering some projects that would improve or add to innovations like the mybus website.
Besides the fact that UW is a pretty good school, it’s not surprising that a Seattle institution would be a leader in creating all these tools. After all, they’re necessary for most people to decipher byzantine bus routing and scheduling. In other cities, people just take the train with its comprehensible route map and short wait times.
Feel free to leave your preferences in the comments, where the students can read them:
A group of students from UW are about to start on some cool transit information technology projects. We have a bunch of ideas on what to do and would like feedback from people on what they would use most or which of our ideas would be the most useful to the general public. So far we have created a sample of one of these ideas, but we would like feedback before we move forward with other projects.
Our ideas include:
1. One Bus AwayIn this application, people will be able to enter an address or click on a google map and be able to see where their nearby bus routes go and what destinations are served along the route. We have created a basic version of this website for destinations in Seattle. Here is a link to that website.
2. Commute CalculatorPeople will be able to enter an address and then our program would tell them the average travel times via car or bus to major employment centers. Users could also just enter their current work address as well. This application could also show the cost to use transit compared to the cost of using a car.
3. Redo BusMonster
In this application we would beef up bus-monster with more graphics and more tools for people to use. We aren’t sure about what features we will add yet, but they will be cool.
4. Advertising with real-time info
Here we would try to start an advertising business with mybus arrival times and ads displayed in prominent storefront windows. We would rent unused space by windows in businesses next to a bus stop. Then we would charge other companies to be able to advertise on our computer screens which would switch between real-time bus arrival times for perhaps 30 seconds and then business ads for 30 seconds.
Thanks in advance for giving us feedback.
I love that bus destination tool!
I love number four. Its a nice way to do offer useful bus data where it is needed. Honestly I’m not too keen on the ads piece of it, but it has to be paid for somehow. One point of clarification I’d like with #4: are the ads specific to the store that the monitor is located in, or are you selling ad space willy nilly?
Another “innovation” I would like to see is a more responsive SMS mybus.org application. The email gateway seems kinda slow at times. Perhaps we could get a real life short code? (Yes, I know this would cost money, but we have one telephone wireless company based in the area, and another with major operations in the area. Engage them and perhaps get them to donate it.)
I think One Bus Away is novel, but you’d definitely want the ability to state a starting address and then it’d give you all the cook destinations on routes near you (maybe w/ riding times too!).
But, in terms of the most useful, I think number four would be the best. I am totally fine with ads because I’d definitely like to see this signs self-supporting rather than depend on future UW grants.
I think Google Transit does a better job of the OneBus option (except it isn’t limited to one bus, therefore making it different).
I’d like to see a variation on the OneBus: A dynamic map that highlights all of the areas you could travel from a given location. Type in an address and choose how far you want to walk, and a map appears showing green areas for areas you can reach via a single bus with frequent service, yellow for moderately frequent, and red for infrequent. The actual bus routes would be lines with bus numbers.
Here is another idea we are playing with. We forgot to add it.
5. Improved Mybus interface and marketing
Mybus is an great tool that is up and running, however very few people even know about it. It is easy to use and available online, over text messaging or over the phone. This project would combine all of the existing mybus information into a unified, user oriented, and marketable package which will then be promoted. It would include information at bus stops, real-time bus arrival information in google maps, a website, and possibly a print map along the lines of this (http://flickr.com/photos/bgtothen/2306344386/sizes/l/)
Number four is a fantastic idea. I wonder if you could sell it to businesses as a way of attracting customers to their store. That might allow you to reduce the dependence on third-party advertising.
I had a number of UI ideas for MyBus. In fact, just yesterday I was thinking of throwing together some kind of demo for the iPod touch. (For one thing, I wanted to see if it was possible to do an offline mode, so I could get some schedules even while outside of WiFi range. The only question was where to get the data.)
how about doing some crazy data mining / plotting the results over a map of king county. generate an animation that is updated in 15 minute increments for all times during the day (averaged over N week days), and plot areas where transit delays (as measured by mybus) crop up. it would be interesting to see if there are patterns in where delays develop and analyze the most consistent trouble spots to see what can be done to improve transit service (the delays may not always be traffic-related).
i also really like putting mybus in storefronts. it’s too bad, aside from the aurora pilot, that metro isn’t bothering to provide riders this service.
someone needs to come up with “The” one stop shop for mobile phone transit info. this content aggregator would at a minimum provide in super-low-bandwidth format:
-schedule info (metro & st +)
-trip planner lite
-mybus (made intelligible)
-wsdot flow map
-wsdot incidents
-seattle 911 info
another idea i’ve been toying with is mashing up seattle 911 data from the cityofseattle.net site (it’s really easy to play with) and mybus – e.g. if there’s a multi-unit-response emergency along the routes serving a given mybus stopID, indicate that there might be a slowdown due to emergency vehicles on the route. this does two things – takes mybus to the next level by providing a potential explanation for why your bus is late, and also provides advanced warning that mybus might not be able to catch (including potential obstacles you might run into after you’ve boarded your bus). at a minimum, mybus could show some sort of icon indicating “possible trouble on route”.
archive the ITS-data feed of all the AVL data from all of the metro buses up on Amazon S3 or something similar to that, so people can write little applications to mine and process that data.
just a few random ideas. i also miss the old mybus web service. i’ve stopped using windows mobile but still would enjoy having web service access to mybus just for my own evil projects.
Andrew – nice ideas. I think we also talked about scraping mybus to find out where buses were late a lot. Also, I was particularly interested in trying to use that data to create 95% reliable times for transit routes, similar to wsdot’s 95% reliable travel times. Also, that would be a cool idea to put that data up on Amazon S3, although that would cost us poor college kids extra money, and I don’t know what KC Metro would think about having that data up there.
Also, mybus still has a webservice available for real-time stop info. Check out their webservice site here.
thanks for the feedback, djstroky :)
amazon’s S3/EC2 stuff is really cheap. if you know anyone or know anyone who knows anyone over at amazon they may know someone who can be convinced that helping out transit in seattle may be a good PR move for amazon (it’s worth a try).
barring that, from my own experience the ITS data feed is actually pretty lightweight so you’d be lucky to feed in more than a gig of data a month ($0.10/gb) – I’d suspect you could pull the project off for less than $8 a month for the quarter, though as you mention i’m not sure what the restrictions are on using the data (it could always be kept private with you guys just exposing various applications on top of it)
that being said, it’s also probably possible to collect and aggregate that data on your own machine though it is always nice to leave the administration to someone else. (n.b. – i don’t work for amazon, just think AWS would be super cool for this type of application)
thanks for the link to the mybus web service. i had been having trouble with it a while back in terms of availability but now will end up playing with it endlessly :)
95% reliable times for routes would be very cool. skewing the notion of reliability a little bit, it would be good to also maybe provide reliability metrics on the schedules as well (for a given timepoint for a given run of a given route, what is its on-time percentage). the data should be all publicly available (and metro i’m sure keeps track of this internally) but damned if the pulic will ever see those numbers even if they’re super useful!
If you’ve got a bike, buses are much more useful: with a bike, you can start or stop your bus trip at a destination much farther from your final starting or ending points, increasing the area you can cover, and often, decreasing travel time because you can pick a distant express bus instead of walking to catch a nearby bus with a lot of stops. You can also sometimes get where you’re going faster by getting off the bus early and biking to an express bus stop that would be too far to get to on foot.
But bus schedule web sites don’t have all the features that would make the bike/bus combination as effective as it could be. Here are my suggestions:
1.) A “I’m on a bike” check box. With this information, a bus trip planner could assume that you can travel further (3 miles?) and faster (10mph?) to stops than you could by walking.
2.) If the “I’m on a bike” check box is ticked, there might be some preference for bus stops near official bike lanes, or trails like the Burke Gilman.
3.) If “I’m on a bike” is checked, the trip planner should also consider the option of getting off in the middle of a bus trip when it would be possible to bike to a faster express bus, or a bus that goes more directly to the desired destination.
4.) Seattle riders have noted that bike racks on buses are increasingly full. If a full bike rack causes you to miss connections, it can really mess up your trip. What will be increasingly needed is some intelligent way of displaying “rack full” contingency routes. A biker could print these out or have them stored on a cell phone.
I would love to see a ‘don’t use this bus’ checkbox. Often there are more options and times that it’s hard to find for some routes.
It seems to me that #1 and #2 could both be implemented in the same interface, or maybe both in a BusMonster interface. I choose my starting location, discover that my work is along one of the routes available. I then select that route and I get a printable schedule for that route (or something that can be easily imported to my phone/PDA) and I get commute calculator results – as times and as dollars saved over car use. For the expense part, would need to be able to input personalized figures like my car’s MPG, yearly insurance and maintenance, and my parking costs.
I also really like number four.
Speaking of BusMonster – it seems broken today; map images are just showing red x’s of broken images?
Number 4 is awesome because it is a great counter-argument to the anti-transit folks who believe public transportation should be ‘profitable’.
And, it’s a highly visible solution that might actually increase ridership.
(I never understood why so few applications are focused on getting new riders as opposed to making life easier for existing riders. New ridership solves all sorts of problems.)