Ahead of this week’s guest-lecturer Adam Millard-Ball, I read his piece, “Parking search caused congestion: Where’s all the fuss?,” in which Millard-Ball and the other authors compare perceived cruising levels versus actual cruising in San Francisco and Ann Arbor. “Cruising,” or the active search for parking, is analyzed in a new light – through GPS traces – and the authors concluded that actual cruising levels in the two cities are less than previous estimates. Millard-Ball et al.’s definition of “cruising” includes definitions of what cruising is not, for example, “non-cruising situation[s],” such as out–and–back errands. I found this interesting, especially after reading the conclusion where this particular type of travel is identified as potentially the dominant reason for excess travel. This methodology for parking analyses is crucial because it allows for parking policies to be individualized, especially in a world where most parking practices follow the leads of other cities, resulting in many poor practices. One way to expand on this research could be to include the actual destination (data permitting) and not just the location of the end trace. If that data was available, I wonder if it would be possible to group the location of destinations from the data set, to separate tourists with multiple destinations from single destination goers. For example, in San Francisco grouping the final destination as Pier 39. In addition to that, what percentage of cruising is attributed to tourists.
Were there any instances where there was interference with the GPS data? Such as if a car entered an underground lot and could no longer transmit the signal?
I did not notice any note of one-way street traffic, could one-way streets increase the cruising distance or time (particularly in San Francisco)? If so, how did you account for the one-way streets?