A GPS Solution by Barry Korengold

In light of the current airport debate, I think it'd be helpful to revisit this proposal for a GPS solution that the SFCDA submitted to airport officials last May.  This will also give some insight into the ideas being considered at the time.  The "two line system" I refer to, was a proposal to have a separate "local" line and a long line. The drivers would choose which line to wait in, with 2 different curb spaces at the terminals.


I think it's time to reconsider this GPS solution.  A system such as this could have feasibly been in place by now.

Barry  Korengold

May 7, 2010

Airport Directors                                                                                                            

Landside Operations Management
San Francisco International Airport

Dear Sirs/Madam,

We’ve considered the various options proposed at the April 28th meeting, including our own proposal of a single line with a minimum $20 or $25 fare from the airport.  We’ve heard from many drivers on the subject, and although we think a $25 minimum would eliminate the angst of getting a “short” ride after waiting long periods, many drivers are afraid of losing too much business and if the minimum was lower, it wouldn’t be enough to make up for the wait.

The SFCDA’s board has concluded that some kind of “short” system is needed to make it worthwhile for cab drivers to spend the time necessary to have enough cabs available at busy times.

We feel there will be problems with a two line system. It would take up more curbspace, and tend to fall apart when it got busy, creating the need for a single line.  Many drivers who waited in the “long” line, would be forced to take local rides.  It would also be very subjective when the starters would start sending cabs from the long line to pick up from the local line.

The New York style system where the starters give the passengers tickets indicating they’re shorts, would be very cumbersome during busy times, requires lots of extra personnel and lends itself to corruption.

We spoke at length with John Wolpert, CEO of UpStart Mobile, a Best Buy spinout, and previously designed systems for IBM.  He feels GPS can be used to verify geographical “local” trips and prevent cheating of the system, but is not quite reliable enough to use for real time verification.  In other words, there are glitches and reasons the signal might be lost temporarily, but that the information could be “cached” and recovered at a later time to help keep drivers honest.

So here’s what we came up with:

Clear geographical parameters would be set for trips considered “local”.  These parameters are subject to discussion, but we suggest Cezar Chavez to the north, Mission street to the west, Daly City, South City, Skyline Blvd, San Mateo, Foster City.  These are all trips of $25 or less.

For efficiency and to avoid backups, only cabs with short or “local” rides would pass by the exit gate.  A sensor would establish what cab was at the gate, the driver would swipe their card and the system would match the driver with the vehicle.  When the driver comes back and goes through the “local” line, he/she swipes his (or her) card as they do now before entering the main lot, and there’d be a record that they went through the local line.

If a driver tried to cheat the system, nothing would happen that shift, but the GPS information would show they went further than the established boundaries or that they turned off their GPS, the computer program would automatically flag these trips and they would later be blocked from the airport for a significant period of time.

If the penalty was stiff enough, drivers wouldn’t try to beat this system.  There should be a slight buffer zone however, so that if a passenger is going just a few blocks beyond the boundaries the driver won’t get penalized.  This could be programmed so that the actual “red flag” zone was a bit beyond the established boundaries.  Egregious violators would be easy to identify and less serious breaches could be admonished on the first violation.

Wolpert feels such a system would be relatively easy and inexpensive to implement, and could reasonably be in place sometime between September and the beginning of the year.

We like this system because it preserves the safety net of the short line, and also prevents corruption on the part of drivers or starters.  It’s not a huge change in procedure, and allows for further testing of a more complete and real time GPS system.

Although problems have been identified with the current time based short system, we don’t feel it’s as bad as portrayed in the media or that there’s an urgent need to get rid of it immediately.  The SF Weekly article was a staged event, with the driver purposely driving crazy fast to make his point.  We feel most experienced cabdrivers are safe and courteous and we should stick with this system until this GPS backed system is in place.

Thank you for considering our ideas, and we urge you to go with this plan.

Barry Korengold
President, San Francisco Cab Drivers Association