Catoe: NextBus beta "doesn't make sense for Metro"

Transit

Catoe: NextBus beta "doesn't make sense for Metro"

by David Alpert • March 6, 2009 12:58 pm

Metro General Manager John Catoe opened his chat today with a statement rejecting the option of opening up a NextBus beta test before the official launch:

Photo by noon3butm3.

There has been a lot of talk lately about Next Bus, which is a good thing because we expect Next Bus to be back up and running this July.
The return of NextBus is highly anticipated by many bus riders, and by many who were able to gain access to an internal test site over the last few months. We have restricted access to that test site, which has disappointed a number of people. Those people who were using the system have reported in blogs that the system was working well for them, and I'm pleased to hear that.
However, I have to take a much wider view of NextBus and the accuracy of its predictions. Launching a "beta" version may make sense for software developers, it doesn't make sense for Metro. Before NextBus is fielded again, I have to be sure that it will work well for all of our bus riders. If we allow access to a test site, then we are in effect launching that site and service. That means we need to be ready to give anyone and everyone using that site our full attention if they have problems and complaints. All the disclaimers in the world won't make any practical difference. If we were to allow access to the site, then, potentially, there would be hundreds of thousands trying to use something that just isn't ready for prime time, yet. That means a flood of complaints, and we take every complaint seriously. If we allow the system to be used before we're confident that its ready, then, although, it may be convenient for some, it just won't work for everyone the way we envision it should. I would also take more resources, both financially and in terms of manpower that we aren't ready to commit yet.
When the system went off-line on October 31, 2007, we estimated that it would take up to 18 months to bring it back. That estimate was fairly accurate. Just be patient a few more months, and the service will be back up. There's a saying "measure twice and cut once." That's what we are doing with NextBus to make sure we have everything in place to meet our goal of 95% accuracy on bus arrival times. This will be a significant improvement, and worth the wait.

Whether Catoe is right or wrong about the beta, we should keep in mind one thing. Metro has been working hard on this system, and we should applaud their efforts. As a commenter from Phoenix pointed out, some transit agencies aren't even working on such systems at all.

A beta test would generate some complaints, yes. But it would also generate useful feedback Metro ought to be interested in hearing. It's always cheaper to make changes early on. Once Metro launches, we'll inevitably discover some small flaws, and have to wait longer to fix those.

It's too bad Metro reflexively clamped down on the system, assuming it would generate complaints instead of waiting to see if it did. Leaving it up would have been a great opportunity to hear feedback while maintaining a plausible deniability. It'd be terrific if they now offered to give passwords to the NextBus test system to all the people who signed the petition and a limited number of others. Still, Catoe does have a point that public access will, to some extent, distract staff. The question is whether they can limit the distraction and gain benefits, such as valuable feedback, that outweigh the distraction.

Ultimately, this system will launch soon. When it does, if it works well, we should thank Catoe and Metro's technology team for making it happen. Their initial reaction may reveal a risk-averse approach to decisionmaking, but at least they aren't entirely innovation-averse.

Speaking of innovation, where's our Google Transit?

17 comments

Tags: real-time bus, technology, WMATA

Zimmerman urges Metro to embrace "beta testing" (Mar 26, 2009)
WMATA removes NextBus public test (Mar 3, 2009)
NextBus accuracy slips, "ghost buses" explained (May 17, 2010)
Nextbus IP, Metro data quality dissuades one developer (Aug 13, 2009)
RAC tonight: service cuts, NextBus (Mar 4, 2009)

David Alpert is the Founder and Editor-in-Chief of Greater Greater Washington and Greater Greater Education. He worked as a Product Manager for Google for six years and has lived in the Boston, San Francisco, and New York metro areas in addition to Washington, DC. He loves the area which is, in many ways, greater than those others, and wants to see it become even greater.

Comments
Add a comment »

I asked Mr. Catoe about Google Transit on the lunch time chat. I stressed that it would increase off-peak riders when there is excess capacity. He decided not to answer my question.

by Erik on Mar 6, 2009 1:15 pm • link • report

I think the bigger concern is the generation of invoices from the service provider. The key is seeing how their contract is written.

by RJ on Mar 6, 2009 1:29 pm • link • report

I can understand why Mr. Catoe might be risk-averse. Not everyone is as WMATA-friendly as the transit advocates on this blog.

Reading the comments section on any Post article about anything Metro or WMATA related would be enough to make me risk-averse. I don't know where some of the miserable trolls on those comment pages come from.

by Cavan on Mar 6, 2009 2:17 pm • link • report

On Google Transit, I don't see why everyone expects WMATA to pay its staff to reformat its bus and train schedule data and then give it -- donate it, essentially -- to Google. Last time I checked, Google wasn't a public service company. It makes money off of what it does, and it makes perfect sense for Metro to expect some sort of compensation for its data and the staff time necessary to package it to Google's liking.

And I find it rather dubious for a former Google employee, who now sits on Metro's Riders' Advisory Council, to continue to push Metro to join a service provided by his former employer.

by Dustin on Mar 6, 2009 2:52 pm • link • report

Dustin: They've already reformatted it. The work is done. We're pushing for them to release it to everyone, not just Google. Michael Perkins was the one who initiated the push on this site. And they should do it because it'll make riding Metro easier for riders, along with bringing in more revenue to Metro from the higher ridership.

by David Alpert on Mar 6, 2009 2:56 pm • link • report

I don't see what the problem is, they're pretty good at ignoring most comments and complaints, why should it be so hard to ignore comments and complaints from NextBus users?

The thing that is the most infuriating about this whole thing is the lack of trust. There is no transparency. There is no sense of what they've been doing since 2007. I have no idea what they still need to do. Nothing. Absent any transparency, all we have is a sense that they set a date that they felt would make their jobs easy and are doing as little as possible to move that date up.

Real time bus arrival information is such a revolutionary change that Metro should have been moving heaven and earth to get it done right and fast. They shouldn't have wasted time or money on the website, which offers hardly any increased benefits to riders.

Fundamentally they are making decisions based on fear, not based on the interests of the riders or with forward-looking visions. That is why they are terrible, and that is why Catoe must go.

by Reid on Mar 6, 2009 3:39 pm • link • report

Oh and notice he doesn't say "June or July", now it's just "July". So don't expect this till Fall at the earliest.

by Reid on Mar 6, 2009 3:41 pm • link • report

Not sure if this came up in a previous NextBus thread, but NextBus does have one serious limitation at the end of the service day: It tells you about the next _arriving_ bus, not the next _departing_ one. For example, I waited quite a long while for a Fairfax City CUE bus at Vienna Metro one evening instead of making other arrangements, because the NextBus display in the station told me a CUE bus was on its way. The bus arrived, disgorged itself of passengers, then took itself out of service. Grph. This is the sort of thing that dissuades occasional riders from considering transit.

Now I feel I have to consult both the NextBus display and a paper bus schedule. Ideally, any service such as this would make it unnecessary to consult paper schedules.

If this is the sort of thing WMATA is ironing out, then bully for them.

by c5karl on Mar 6, 2009 3:46 pm • link • report

What's really sad is that the unpublished beta wasn't hurting anyone, but was really helping the large number of people who knew about it. As this drama played out over the last 10 days, I was willing to hope that maybe this tempest could make an impact on WMATA, so others could share the beta. It's now clear, as I feared, that our pleas fell on deaf ears, and no one will get to use NextBus before WMATA decides it's perfect enough to launch.

Maybe now the bloggers who not only wrote about the beta but also asked WMATA about it -- Dave, Andrew, and Sommer -- will understand why not everyone thought it was such a good idea.

As for Catoe's comments, I can barely restrain my contempt. At best, they underscore WMATA's basic failure to understand the development and deployment of technology: "We can't release anything until it is perfect, because one mis-reported data point is as important as 100,000 accurrate ones." Is it really more important to have no complaints on NextBus than myriad complaints about the bus schedule? (Which surely under-represents the public's view, because really who bothers?)

At worst, the proffered excuse is a smokescreen and WMATA is deliberately lying about the real reason -- probably budgets and contracts. I personally find it utterly mendacious to hold up the sanctity of their complaint-handling process as a reason for anything. As if the problems of individual, inconvenienced riders have ever had the "full and complete" attention of that organization!

As for the 18-month estimate, it's not "fairly accurrate" until we see the official launch. Here's a countdown clock to August 1, 2009: http://tinyurl.com/akm68w Shall we place bets on when NextBus will actually go-live?

by John on Mar 6, 2009 3:53 pm • link • report

You know, there is a way to obtain the agreements between WMATA and Nextbus. It's not quick but you will get a copy of the contract.

I would do it but I already have two active requests in the system right now and I try to limit the amount of requests I make so as not to be a total pest.

Anyone who wants to make the request, email me for tips.

by Michael Perkins on Mar 6, 2009 4:33 pm • link • report

Its already been mentioned that the NextBus sign at Vienna will lie to you at the end of rush hour (6:15 and later) about the next "arriving" bus. The system announces "arrivals" for buses that are in fact scheduled to go out of service.

I was caught in this SNAFU on Monday and just hopped on the Mason shuttle instead of waiting 28 minutes for the next CUE bus that would be picking up passengers to arrive at Vienna.

This is one flaw that should be addressed before NextBus is launched in DC and should be fixed immediately in Fairfax.

by Whiskey Bacon on Mar 6, 2009 5:47 pm • link • report

I agree with the criticism of Catoe's response. It seems to me a bizarre, and quite flawed line of thinking, for many reasons:

- a test site is fundamentally, and obviously, not the same thing as launching a full service site

- explaining that to the public is not a difficult thing to do

- assuming that the public will not understand reflects (as someone above said) a deep distrust of the public

- taking "every complaint seriously" first off, seems just plain inaccurate if you have ever tried to actually talk to someone at WMATA

- even if it were true, taking "every complaint seriously" is a choice, and one that a manager can change. Saying "we do not read or respond to complaints about our Test Site, because it is only a Test Site" is a pretty justifiable choice, it seems to me

- lastly, "measure twice, cut once" shouldn't be the model for designing services that thousands of people will engage with on a daily basis - more like measure, cut, try on, measure again, cut again, try on again, measure again...etc. The idea is that the public can provide incredibly valuable input to guide the development of services as they are being designed and implemented.

There might very well be some legitimate reasons not to have a test site - but the line of thought in this statement isn't one of them!

by Alayna Buckner on Mar 7, 2009 1:43 am • link • report

Re: the last bus arrival/departure issue, I think NextBus can account for that. When you use the interface, there's an option not only for boarding location but also the destination (which is optional). If you picked a destination of any sort, and the next bus arriving was ending for the night, I think it would exclude that bus.

by Joey on Mar 7, 2009 2:33 am • link • report

Speaking of responding to complaints: I say a good and proper response to Catoe's ridiculous statement is to call in and complain every single time the paper bus schedule is off (i.e. every single time you've ever ridden the bus).

If they take every complaint seriously, they'll clearly take down the paper schedules right?

Here's the phone number:

202-637-1328

Here's the online complaint form:

http://www.wmata.com/about_metro/contact_us/ridercomment.cfm

Catoe promises that Metro will take each and every complaint seriously. I say we flood him with complaints about how shitty the situation is right now.

I suggest a routine complaint something like this:

"I'm standing at this stop and the schedule says there should be a bus here now, but it's not. I would like this matter addressed seriously."

We can't give up on this. Remember Google Transit. It took about three tries before they finally admitted the real reason why they were saying no. I think we're still on fake reason one with Nextbus. We may not get them to break, but we can't let them get away with this deceitful, disrespectful, and illogical answer.

by Reid on Mar 7, 2009 2:42 pm • link • report

The Catoe remarks smacks of typical bureaucratic thinking, worrying more about unlikely liabilities than the expansive opportunities.

The problem is not technical, it's psychological; WMATA just doesn't "get" it.

Let's compare this attitude with private airlines that must actually compete with each other for customers. They have been posting live tracking data for years.

by Capitol Dome on Mar 7, 2009 3:43 pm • link • report

NextBus has been available out here in the Bay Area for a few years. In my experience (w/ AC Transit), it is fairly accurate. However, across the Bay, many Muni riders complain of problems. Sometimes buses would disappear off the predictions, or they would dramatically change. I've seen it go from 1 to 2 to Arriving to 7 minutes before. Sometimes it will jump from 8 minutes to 2 minutes in the matter of 60 seconds, giving you little time to jump out of your door and to the stop. Another thing I would like to see is bus numbers associated with each prediction and/or current load on that bus. So rather than seeing:

Next Route 51 vehicles in:

2 min

3 min

I could see...

#5307 2min 95%

#5308 3min 25%

Here is a Q&A Muni Diaries did with one of the NextBus engineers...

http://www.munidiaries.com/2009/03/06/md-exlusive-qa-with-michael-smith-of-nextbus/

Currently, CTA is the only agency I know of that puts bus #'s on their bus tracking site. http://www.ctabustracker.com

by Stephen on Mar 9, 2009 8:29 pm • link • report

If we allow access to a test site, then we are in effect launching that site and service. That means we need to be ready to give anyone and everyone using that site our full attention if they have problems and complaints.

Truly absurd. If we'd had this attitude, he'd still be waiting for the opportunity to try out that new "Google" thing people have heard about.

It had been up for months. Did they really have hundreds of thousands of people demanding their attention every minute of the day? I think they didn't even know.

by David desJardins on Mar 12, 2009 2:31 pm • link • report