No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Parking-Cruising Caused Congestion
... Another method frequently adopted to recognize cruising in GPS traces is to analyze the amount of excessive routing, i.e. the extra distance traveled compared to the optimal route to the parking place. Weinberger et al. [26] proposed a heuristic to try to understand when the user started looking for parking, and applied machine learning to classify whether ended trips contain a search or not, basing the predictions on the ratio of the actual paths to the shortest route. Even Millard-Bell et al. [15] considered the last portion of recorded trips and computed the difference between the actual distance traveled and the respective shortest path distance. ...
... For this reason, previous studies also propose a personal evaluation metric based on empirical trials, and there is no fixed common guideline. Montini et al. [16] proposed a range of 800 meters, supporting the choice as an overestimation of the previous boundaries of parking search of 350m found by Weinberger et al. [26]. Millard et al. [15] did not need a radius estimation, as they performed the cruising detection based on the ratio between excessive travel and the shortest path to the destination. ...
Interacting with a smart parking system to find a parking spot might be tedious and unsafe if performed while driving. We present a sys- tem based on a Boosted Tree classifier that runs on the smartphone and automatically detects when the driver is cruising for parking. The system does not require direct intervention from the driver and is based on the analysis of context data. The classifier was trained and tested on real data (615 car trips) collected by 9 test users. With this research, we contribute (i) by providing a literature review on cruising detection, (ii) by proposing an approach to model cruising behavior, and (iii) by describing the design, training, and testing of the classifier and discussing its results. In the long term, our work aims to improve user experience and safety in car-related contexts by relying on human-centered features that implicitly understand users’ behavior and anticipate their needs.
... For stage 3, we associated parking time with residential density and used different values for inner and outer zones of the study area, as done by Salonen and Toivonen (2013). We derived the time spent walking to and from the car from empirical studies (Weinberger, Millard-Ball, and Hampshire 2016). We calculated stage 2 using the r5 routing engine that relies on our updated road network. ...
... In contrast to the previous estimates of the share of traffic that is cruising on specific road segments, Weinberger et al. (2016) analysed a large corpus of GPS data from vehicles to estimate the share of all trips that included cruising at the destination in the cities of San Francisco, CA and Ann Arbor, MI. They developed machine learning algorithms to identify trips that contained excessive cruising. ...
We propose a new way to measure the share of traffic that is cruising for parking: observe how many cars pass a newly vacated space before a driver parks in it. This statistical method provides a quick, cheap and approximate way to estimate what share of traffic is cruising. Using 876 observations of newly vacated on-street parking spaces in central Stuttgart, we estimated that 15 per cent of the traffic was cruising for parking.
This article analyses the impact that different parking management policies may have on public roads. Policies were simulated using a new parking model based on two sub models: choice of parking place and search for parking place. The model considers curb traffic and was implemented into a traditional microsimulation traffic software. The parameters for the sub models were estimated from data collected in the city centre of Santander (Spain) and from a stated preferences survey asked to users of parking spaces. The model for testing policies was run on Aimsun simulation software creating a personalised API programmed using Python 3.7. The proposed model was able to dynamically simulate various policies based on charging for on-street parking spaces with fare updates at short time intervals of between 5 and 15 min. A sensitivity analysis was performed on different fare scenarios and considering different levels of information available to the users. As a result, this work demonstrates some benefits of dynamic fares such as reducing searching time, curb induced traffic and emissions as well as a new modal redistribution of parking choice between off-street and on-street supply. On the contrary, dynamic fares implied that users needed to spend a bit more time from their parking location to their destinations.
ResearchGate has not been able to resolve any references for this publication.