Wednesday, February 11, 2015

The MBTA Must Embrace Data!

As you may have heard, we’ve had a bit of snow in the Boston area recently.  Two storms, one the beginning of last week and one which just ended yesterday, each dumped close to a meter of snow in the area.  The two storms each had different profiles: last week’s storm featured rapid snowfall and furious winds, with the snow falling over a 24-36 hour period.  The more recent storm started on Friday afternoon, ended on Tuesday morning, with a steady fall of lazy snowflakes.  Last week a hare, this week a tortoise.  But both weeks, a paralyzed Boston from a transportation standpoint, with the MBTA mass transit system performing dismally.

Unfortunately, the main response to that failure has been a lot of political theater. GM Beverly Scott gave a press conference yesterdaythat featured the usual refrain: the system features antiquated equipment, our crews are working hard, nobody could deal with this.  In other words, a string of unquantifiable and unactionable clichés.  There's already an unhelpful murmur in the press that Scott might be fired, which would seem little fix but mostly fodder for more column inches of newspaper opinion (such as this and this)
Despite Boston priding itself on being a Hub of big data and science (and a new governor who claims to be data-driven), the T’s management never gives the impression of being interested in data. But that’s where the conversation desperately needs to move away from wishing the snow were gone (alas, Ms. Scott did suggest that solution) to details, details and more details.  We need the T to be providing them – and the Boston Globe and other media to be demanding them.

For example, one of the trouble points on the rail systems have been frozen switches.  Some switches are defended by heaters (if you see the rails smoking, that is generally not trouble but small, contained fires keeping the switches free of ice); many are not.  But which switches are undefended, and which ones are critical?  How much does it cost to install a switch heater? What would the total tab be across the system?

Three of the heavy rail (“subway”) lines are powered by third rail, and a Globe article this weekend  outlined that older DC motors are much more likely to fail than newer AC motors.  One line (Blue) has all new cars with AC motors, while the other two lines are either all DC (Orange) or a mix (Red). It was a good article, with a nice plot of car ages on the three lines -- though, alas, not clearly indicating the DC vs. AC motor issue the article was centered on and also not quite explaining why AC motors are more robust.   The T has said they’ve burned through (and unfortunately, failure for the DC motors is destructive) their entire stock of replacement motors during these two storms.  What is their stocking level and how long do they usually last?

Truly, it is sad that if I want to know how long out of action a member of the Red Sox or Patriots will be, that is easy to find by Google or will be reported in the Globe.  But if I want to find out how many diesel commuter rail locomotives are sidelined and when each one is expected back, I should give up now.

Even discussing the snowfall has lacked data.  Officials at the T and elsewhere have described these recent storms as “extraordinary”.  They’re clearly  large, but truly outliers?  I doubt it; this is New England, and most winters feature a storm or two of similar magnitude.  In any case, it’s still the wrong question.  What needs to be asked of the T is what magnitude storm can they deal with? What fraction of storms over the past decade or so are outside those bounds? The T had to borrow track clearing equipment from the New York City system (which was crazy enough to lend it); how much did this improve their abilities and what would it cost to have these permanently?

Today, the various rail lines are running, albeit on reduced schedules.  Were these reduced schedules simply thrown together, or are these long-standing contingency plans?  How much wargaming of contingencies does the MBTA run?  Does the T, or the commuter rail operator Keolis, explore extreme strategies to keep portions of the system running?  For example, essentially four lines run out of North Station (one splits many miles away_ and the critical drawbridge has four tracks; could the reliance on switches be reduced by treating them as 4 one-track lines?  Could a fraction of the Orange, Red or Blue lines be kept running by restricting running only to focus on the tracks that are in subway tunnels?  If the T does think out-of-the-box and find the solutions don’t work, they need to show their work!

Keolis, operating the commuter rail under contract, actually must pay penalties for poor service.  What are the magnitude of these penalties vs. the amount Keolis or the MBTA are spending on reliability improvements?  Do the penalties get earmarked for improvements or just disappear into the maws of the MBTA?

A non-solution is to just ignore the problem.  The MBTA provides a critical component of what makes Boston a productive city, eliminating tens of thousands of daily car trips and the need to park all those cars. Reliable transit is an asset which enhances Boston’s marketability as a place to work and visit.  If the slide in T performance is to be halted, something more than just “give us more money” needs to happen.  We are a big data town; we need a big data transit system.


Jonathan Badger said...

Although those of us who live/have lived in the Washington DC area have to feel a bit of schadenfreude over the MBTA's problems because typically it's the DC Metro that's the mass-transit system with issues...

AMac said...

> The T had to borrow track clearing equipment from the New York City system (which was crazy enough to lend it)...

Not crazy -- you're almost talking about an old-fashioned but good idea. Mutual Aid, most famously used by fire departments and electric utilities.

Utilities from different geographies (unlikely to be laid low by the same storm) agree ahead of time to pool crewed trucks to help restore power. These days, the loaner trucks are often dispatched and pre-positioned on the basis of the weather forecast.

Does the T have arrangements like this with sister organizations? If not, why not?