Technology Lock-In with the DC Metro


SmarTrip cardI think that sometimes when technologists make the case for open standards it can seem like a purely theoretical exercise. For most people the downsides to publishing a document as, say, an MS Word file aren’t readily apparent. Every computer they’ve used has had a Windows license built into its price. They’ve never had a reason to learn how to manipulate text programmatically. Everyone else with whom they exchange files has Word, and the program is pretty well-designed for most office work use cases*. The dire warnings issued by developers just don’t seem plausible.

So it’s worth taking a second to note an example of these problems happening in a different arena. Here in DC our primary transit agency, WMATA, issues an RFID card called the SmarTrip which works with nearly all of the area’s various transit systems. It’s quite handy: you don’t have to take it out of your wallet to use it, the balance is supposedly loss- and theft-proof, and it automates things like bus transfers.

Unfortunately, this morning brought news that the SmarTrip has to be replaced. Why? Well, the vendor that our transit planners bought it from has gone out of business is ceasing to support the card, and they’re pulling SmarTrip into oblivion with them is ceasing to support SmarTrip, and no one else can take their place: the card incorporates proprietary technology, so it’s impossible to find a new supplier. WMATA has a stockpile of cards that’ll last about two years, but after that it’ll have to start using a new solution.

Will the readers on every bus and subway turnstile have to be replaced? Will riders have to pay for new cards? How will balance transfers from the old cards to the new ones work, and how much money will be lost along the way? Whatever the specifics turn out to be, it seems safe to say that this will cost WMATA and its riders attention and money — and both are in short supply right now.

It didn’t have to be this way. There are open standards for RFID cards. A little more care during the procurement process could have avoided this mess. Unfortunately, it seems that adopting open standards wasn’t a priority.

In the software example, things are admittedly a little different. Microsoft’s not on the brink of bankruptcy, and while Word’s formats are far from ideal, they don’t have quite the same IP encumbrances that the SmarTrip does. Still, the principle holds: it’s wise to choose openness today to avoid headaches tomorrow. It’s a lesson that’s well known by the people reading this blog, but one that we need to keep teaching.

  • As a sidenote: I’m convinced that if someone built a lightweight text editor that supported Markdown, autosave and had good Track Changes support, every freelance writer and their editors would buy a copy. Don’t say OpenOffice.


A few people in the transit technology community have taken issue with this post; let me address a few of their objections. First and foremost: I screwed up when I said the SmarTrip vendor is going out of business. I got that wrong — in fact, the problem is that they’re discontinuing the technology that SmarTrip is based on. But they’ll still be in business.

I find the other objections raised less compelling — here’s a post covering many of them. I think folks are mistaking the forest for the trees: the idea that openness is the way to avoid lock-in problems should not be controversial. Nor, as folks have pointed out in comments, is it a very good idea to allow a procurement process to result in ongoing reliance on a single vendor. These are the central points I was trying to make, and I don’t think they’re in question.

I suspect that people are understandably sensitive to what they view as yet another shot at Metro, an agency that gets more than its fair share of criticism. I’m sensitive to that, too. But this isn’t about beating up WMATA; it’s about illustrating a point.

On the technical matter of whether a good open standard existed when WMATA launched SmarTrip: it’s pretty tough to trace back the chronology of the relevant ISO standards, and I’m grateful to those that have shed light on the issue since my post went up. It sounds like the first relevant ISO offering came in 2000; SmarTrip launched in 1999. This means there was probably a working group that could’ve been engaged with, but it’s hard to get too irate about WMATA’s failure to find it.

On the other hand, it’s not as if SmarTrip has been set in stone since ’99. There have been various upgrades and revisions to the system and contract since then. This problem could have been addressed prior to the current budget crisis and technology phase-out deadline, but it wasn’t (in fact, it sounds like the contract wasn’t managed very well at all).

Besides, “open” doesn’t necessarily mean “ISO standard”. Insisting on the right to fully use the design and software you’ve paid for is a pretty obvious thing to do. Or you could write into the contract a requirement that the vendor open source the relevant assets if they become deprecated. There are options. WMATA’s Nextbus contract is a decent example of what’s possible: that vendor is also selling a proprietary technology. But the question of lock-in is less vexing, thanks to a clear understanding of who the data and technology belong to. Nextbus’s prediction algorithms will remain their own, but WMATA’s forthcoming bus location API — built on modular, open technologies — should make it possible for others to step into the gap if Nextbus disappears.

But now we’re clearly in the weeds. And as I said, I’m not particularly interested in assigning blame in this situation — I just want organizations to make smarter decisions about the standards they use. Hopefully we can all agree about that.