Rachel Weinberger’s research while affiliated with University of Pennsylvania and other places

We analyze GPS traces of 5,316 trips that are cruising for parking in San Francisco and Ann Arbor, and use cluster analysis to develop a typology of five distinct types of search strategy. Our most striking finding is that most cruising trips do not involve circling. Partly because most drivers are able to find a space relatively quickly, a more typical cruising pattern involves just a few turns. While drivers often perceive that cruising times are long, most cruising trips appear to be less dramatic; repeated circling is the exception rather than the norm.

This paper presents a method for determining parking search behavior using GPS traces. The research takes advantage of a GPS based household travel survey, an extensive dataset of GPS with video, and a commercially purchased set of trip segments. Strategies for data cleaning, matching traces to digitized networks, assessing the probability that a trace is of good quality, and strategies for determining whether or not a trip involves excess travel due to parking search are described. We define and operationalize two definitions of excess search – popularly known as cruising. Our results suggest that cruising in San Francisco, CA and Ann Arbor, Michigan is acute in some locations but overall experienced in less than 5–6% of vehicle trips, and that it accounts for less than 1% of vehicle travel in these cities–considerably less than in previous estimates.

Cruising for parking has long been perceived as a major source of congestion and emissions in urban areas, but recent empirical work suggests that parking may not be as onerous as folklore suggests, and that the amount of vehicle travel attributable to cruising is minimal. In this paper, we reconcile these perspectives through a dynamic programming model of parking search, and empirical insights from a large-scale GPS dataset in San Francisco and the California Household Travel Survey. We first draw a conceptual distinction between parking search, the time between the driver’s decision to park and when a parking space is taken; and cruising, defined as excess vehicle travel due to parking search. In places with little or no through traffic, up to half of traffic can be searching for parking, but cruising can be zero. We then operationalize this distinction through a dynamic programming model. The model predicts that when parking is perceived to be scarce, drivers are more willing to take a convenient available space, even if it is some distance from their destination. Counter-intuitively, scarce parking can even suppress vehicle travel as perceived parking scarcity leads drivers to stop short of their destinations and accept a longer walk. Empirical data from California indicate that neighborhood density (a proxy for parking availability) has little impact on cruising for parking, but increases walk distances from parking locations to final destinations. We conclude that cruising for parking is self-regulating, and that in certain circumstances parking scarcity can even reduce vehicle travel.

Automobile ownership and mode choice are typically analyzed as functions of individual characteristics and features of the transportation system. Individual characteristics reflect preferences; transportation characteristics capture context. Increasing interest has been devoted to understanding to what extent preferences are learned. Using the Census Bureau’s PUMS data, we develop a multinomial probit model accounting for the endogeneity of automobile ownership to investigate whether commuting mode choices are influenced by past exposure. We proxy exposure by previous residence, differentiating people who previously lived in “transit-robust” metropolitan areas, metropolitan areas, and non-metropolitan areas. Consistent with previous work, we find that, among recent movers, past residence has an impact on decisions with respect to auto-ownership which we understand as habit formation. Mode choice decisions for recent movers, on the other hand, appear to be made more on prevailing conditions and modal trade-offs faced in the present. The results suggest that for people moving to transit robust cities (1) walking and bicycling are linked to past residence; (2) automobile ownership is inversely correlated with the utility of walking and both decisions are made simultaneously; (3) though automobile ownership is inversely correlated with utility for transit use, there is no endogeneity in the two decisions; (4) while finding evidence that habits formed in the previous residency with respect to automobile ownership, these effects on mode choice appear to be small in magnitude. Mode choice decisions by recent movers are predominately made based on the environment rather than being influenced by past experience.

Global Positioning System (GPS) data have become ubiquitous in many areas of transportation planning and research. The usefulness of GPS data often depends on the points being matched to the true sequence of edges on the underlying street network – a process known as ‘map matching.’ This paper presents a new map-matching algorithm that is designed for use with poor-quality GPS traces in urban environments, where drivers may circle for parking and GPS quality may be affected by underground parking and tall buildings. The paper is accompanied by open-source Python code that is designed to work with a PostGIS spatial database. In a test dataset that includes many poor-quality traces, our new algorithm accurately matches about one-third more traces than a widely available alternative. Our algorithm also provides a ‘match score’ that evaluates the likelihood that the match for an individual trace is correct, reducing the need for manual inspection.

The number of cars searching for parking, also known as “cruising,” is a risk factor linked to increased pollution and congestion and decreased road safety. Although the detrimental effects of cruising are known, the actual amount of cruising is unknown. A novel video data set of naturalistic driving is shown to provide reliable estimates of cruising behavior. The distribution of search start times, search distances, and search times is characterized. Cruising behavior variation between 109 different drivers is also reported on in the study, located in southeast Michigan. It was found that 30% of the drivers generated more than 70% of the meters cruised. This finding suggests that the search strategies of a few drivers disproportionately affect the many. These results facilitate the estimation of the number of vehicles searching for parking and the amount of pollution generated by cruising drivers. Researchers may also use these results to develop more realistic models of parking search and parking interv...