Category Archives: Uncategorized


It’s been a while. I had some time in the last week and created the attached BART map. Only need $10 billion or so to make it a reality.

New BART Map


Timed Transfers: A Potential Strategy to Maximize Transit Effectiveness

This summer, my internship has enabled me to make great use of the Bay Area Rapid Transit (BART) commuter rail system.  BART maintains a guaranteed timed transfer system between two of its intersecting lines: train arrivals are synchronized to let passengers change trains almost seamlessly.  This not only reduces BART’s operating costs, but also increases convenience for its customers.  Since large-scale infrastructure improvements will continue to cost hundreds of millions of dollars for the foreseeable future and operating costs will continue to climb, implementing a system of timed transfers is a strategy that agencies can pursue to make the most out of their existing facilities and limited budgets.

The clearest customer benefit of guaranteed timed transfers is reliability.  Knowing that your connecting bus or train will be there when you disembark your original transit vehicle provides critical peace of mind, a phrase seldom associated with today’s transportation systems.  As a BART rider, I know that if I need to transfer, it doesn’t matter if my train is running a minute or two late; BART will take me from Berkeley to San Francisco in under 25 minutes.  This allows me to better plan my daily schedule and avoid giving my friends super conservative predictions like, “I’ll be there sometime between 8 and 9.”

Timed transfers also eliminate the “agitated dash” of inter-line travel.  As a rider of the intersecting Green and Silver lines in Los Angeles, I’ve frequently found myself dangerously sprinting down the two story staircase at the Harbor Freeway Transit Station, praying that I don’t see the disappearing taillights of the Silver Line when I reach its platform.  The MacArthur and 19th Street timed transfers on BART, on the other hand, are virtually stress free.  I know my connecting train operator knows that I’m walking across the platform and that he should leave the train doors open for as long as I need (though I’m convinced some BART operators like to close the door on the slowest transferring passenger in a Darwinian-like reminder to hurry up!)

While timed transfers at line termini may be easier to implement (take for example, an empty bus waiting at a commuter rail terminal for an arriving train), it’s transfers midline that have the best potential benefits.  Midline transfers allows passengers on both intersecting lines to utilize the timed transfer as opposed to only one line worth of passengers in transfers placed at line termini.  This means many more passengers (as well as many more origin-destination groups) can capitalize on the timed transfer.

From an agency standpoint, timed transfers can save money.  In BART’s case, in the southbound morning commute, there are two typical origins (west Contra Costa County and East Contra Costa County) and two typical destinations (San Francisco/Peninsula and Alameda County).  Instead of needing four trains to serve these four O-D pairs, the timed transfer lets BART only use two.  This means two less train operators, less rolling stock, and additional line capacity for other services.

In order to maximize the reliability gains from schedule coordination, transit systems require certain characteristics.  First, the individual lines need to be relatively reliable themselves because timed transfers mean that delays in one line affect the other coordinated line.  Second, there should be strong enough demand for all four origin-destination pairs; otherwise, you will be unnecessarily delaying passengers that don’t need to transfer between lines.  Finally, there should be minimal walking distance between transfer points.  This allows agencies to not only standardize the times that operators need to wait for transferring passengers, but also keep passengers from worrying that they may miss their connecting service.  Timed transfers aren’t feasible on all systems, but they can be a great strategy to increase reliability for customers and reduce operational costs for cash-strapped agencies.

Sam Levy is President of the USC Institute of Transportation Engineers