Baltimore civic hackers freed bus data MTA priced at $600K — it was already streaming

In February 2015 the Maryland Transit Administration launched a basic web app for tracking Baltimore buses — while telling developers that opening the data to third-party apps would cost roughly $600,000: its 25-year-old CAD/AVL system did not emit the GTFS format transit apps need. The agency's own app was a byzantine web page without geolocation, and the data behind it stayed locked away from Google Maps and the “app economy” the agency's press releases liked to invoke.

Within days, geo-data developer and open-data advocate Chris Whong opened the web app's network traffic and found it polling MTA servers every 10 seconds for JSON — encoded in GTFS, the exact format developers were told was too costly to produce. A few hours of trial and error confirmed the feed; with a small team of civic hackers he assembled a framework that let any application pull the data for free, demonstrated it with a live tracking site on Heroku, and posted everything on GitHub.

Adoption was immediate. Montreal developer Transit App used the feed to launch Baltimore real-time tracking “with just an afternoon of work — $600,000 under budget,” generating its own predicted bus arrivals and comparing them against the MTA's tracker to verify accuracy.

The feed the agency called too expensive to produce was already live inside its own web app, polling JSON every 10 seconds — read the network traffic and republish it.

The costly part — formatting CAD/AVL output — had already been solved once for the agency's own web app, so a machine-readable copy existed that any browser could read.

Publishing the scraped feed as open-source infrastructure turned one developer's hack into something commercial developers could adopt the same week.

The agency's cost estimate was never tested against the cheap path: its own front door.

Before accepting a six-figure estimate for opening data, check whether the data already flows to a system you can reach; the expensive part is often already built for the agency's own use.

The MTA responded to Transit App’s post by explaining the $600,000 figure reflected converting its CAD/AVL data to real-time GTFS before the web interface existed, saying a brand-new real-time GTFS system was already in the works, and announcing engineers had spent weeks finding a way to move the data into a secure environment where third-party developers could safely use it — adding, “civic hackers have helped us speed this announcement up.” Ars noted the underlying data stayed limited: buses report over radio from GPS, so positions drop out until MTA fixed the feed.