You’ve heard of the Baltimore bustracker by now, even if you don’t ride the bus.
Launched about two weeks ago, MyBustracker is an interface the Maryland Transit Administration designed to help riders figure out when the next bus is coming.
Generously speaking, it’s a costly, clunky, retrograde disappointment.
Some in the tech community have been less generous.
“Based on what I have seen so far, they seemed to have hired the finest transit technicians from 1989 Albania. This system might win 3rd place in a 5th grade science fair,” one of them wrote in The Brew, a comment highlighted in the Ars Technica coverage of the bustracker debacle.
Bottom line: it took three years to develop, cost $2.7 million and has real time data only about a third of the time.
To call mybustracker.mta.maryland.gov a multimillion dollar online version of the printed bus schedule, wouldn’t be a stretch.
Questions and Answers
But when MyBustracker debuted, riders, transit geeks and even The Brew were cautiously optimistic that it could improve the transit experience for Baltimore’s ignored and often irate bus riders. Is there an explanation for the gaps and glitches? It’s in beta mode, is there more to come soon?
We sent the MTA a lot of questions and posted their replies online as a Q&A.
Their answers painted a picture of a backward transit agency that wasn’t going to have any sort of real-time cellular system for three years.
The Q&A generated a wave of complaints from the tech and transit community, many voiced on the Facebook group, Baltimore Transit.
Boiled down, the general consensus on MyBustracker after its first few hours was:
It was awful, and the civic hacking community could do better – they could at least put the data in app form so it would be easier to use on your phone, while you’re waiting for the bus.
Enter Chris Whong, a Brooklyn, N.Y., technologist and civic hacker who lived in Baltimore for ten years. Whong heard about the disappointing tracker on Baltimore Transit, a group that he founded.
“There were a lot of complaints, especially about there not being a developer API, ” Whong said, in an interview with The Brew.
Liberating the Data
Whong, who works in urban planning, hardly ever rode the MTA buses while he lived here (he did ride the Circulator) but decided to take a crack at making developer API available.
“I just started tinkering with the MTA’s app to see how it got its real-time data, and whether that data could be transformed and consumed in other ways, “he said. “ First we figured out how to query the MTA’s servers for all buses that are reporting real-time locations, and we just threw them on a map.”
He liberated the data in about a week, posting his project on Github, a code sharing host. The data is open source.
The map by Whong (and fellow developers Elliott Plack and Shea Frederick) is a “proof of concept” which has undergone several iterations.
The map page auto-refreshes every 30 seconds. Users can click on bus routes and see how many (if any) are reporting GPS real-time data. It varies according to time of day, peaking at about 169 during morning rush hour, and close to zero in the middle of the night.
“We Saved the State $600,000″
On Monday, TransitApp, a Montreal-based developer, announced that they had saved the state $600,000, by making MTA’s bustracker web site data accessible as an app using Whong’s data.
The $600,000 figure comes from The Brew Q&A, where MTA said it would have cost them approximately $600,000 more to be able to format data from their 25-yr-old CAD/AVL system into GTFS (bus schedule data) for use by outside developers.
Whong had done it for free.
Jake Sion, TransitApp marketing director, speaking to The Brew this week, said his company reached out to the MTA not long after MyBustracker debuted, but didn’t get a response.
TransitApp already has 20,000 customers in the Baltimore area for other transit information, and had the data been available from MTA, he said, they would have made an app for little or no cost.
“A transit agency’s core competency is providing transit. They are not a tech firm,” Sion said. “It makes sense for them to get out of the way.”
The MTA fired back with their own statement about “My MTA Tracker for Bus,” as they refer to the app in press releases.
In “Did civic hackers save the MTA $600,000 in one day? Not quite but we’re glad they tried,” the agency disputed the notion that the hackers had embarrassed the agency with a quick freebie version, but welcomed the help of “our private sector friends” and, most importantly, promised to make real-time data they need available.
Bringing Real Data to the Table
Whong is glad to have sparked change and careful to not over-promise. He says his map project isn’t terribly useful to bus riders day-to-day, “but it helps us get a sense of how many vehicles are reporting over the course of the day.”
“One goal of the project is to log the real-time reports. . .so we can assess how each route is performing against its schedule, and where/when there are gaps in real-time coverage,” he said.
What Whong is really measuring is the true performance of the MTA buses: how many are really on schedule, how many are not, and how many are AWOL.
“If we can put together a simple visualization or tool that shows gaps in coverage for real-time data, transit advocates in Maryland can bring some real data to the table if they ever need to criticize the MTA about this program,” he says.
The policy implications, for Whong, are much broader.
“This kind of data shows how a poorly-functioning transportation system has negative repercussions throughout the city for economic development, real estate, jobs, etc,” he says.
“I would like to see this lead to real change, meaning a better and more reliable public transportation system in Baltimore.”
MTA to Whong: Let’s talk
So far, no transit equity activists have a expressed a particular interest in the project or the data,Whong said, but now the MTA’s has reached out to him.
“They want to bring us in for a discussion,” Whong says.
What’s more, the MTA has finally agreed to making their real-time tracking data available to developers.
They just won’t commit to a timeline yet.