ParkNet: drive-by sensing of road-side parking statistics

  • Authors:

  • Suhas Mathur;Tong Jin;Nikhil Kasturirangan;Janani Chandrasekaran;Wenzhi Xue;Marco Gruteser;Wade Trappe

  • Affiliations:

  • WINLAB, Rutgers University, North Brunswick, NJ, USA;WINLAB, Rutgers University, North Brunswick, NJ, USA;WINLAB, Rutgers University, North Brunswick, NJ, USA;WINLAB, Rutgers University, North Brunswick, NJ, USA;WINLAB, Rutgers University, North Brunswick, NJ, USA;WINLAB, Rutgers University, North Brunswick, NJ, USA;WINLAB, Rutgers University, North Brunswick, NJ, USA

  • Venue:
  • Proceedings of the 8th international conference on Mobile systems, applications, and services
  • Year:
  • 2010

Quantified Score

Hi-index 0.01

Visualization

Abstract

Urban street-parking availability statistics are challenging to obtain in real-time but would greatly benefit society by reducing traffic congestion. In this paper we present the design, implementation and evaluation of ParkNet, a mobile system comprising vehicles that collect parking space occupancy information while driving by. Each ParkNet vehicle is equipped with a GPS receiver and a passenger-side-facing ultrasonic range-finder to determine parking spot occupancy. The data is aggregated at a central server, which builds a real-time map of parking availability and could provide this information to clients that query the system in search of parking. Creating a spot-accurate map of parking availability challenges GPS location accuracy limits. To address this need, we have devised an environmental fingerprinting approach to achieve improved location accuracy. Based on 500 miles of road-side parking data collected over 2 months, we found that parking spot counts are 95% accurate and occupancy maps can achieve over 90% accuracy. Finally, we quantify the amount of sensors needed to provide adequate coverage in a city. Using extensive GPS traces from over 500 San Francisco taxicabs, we show that if ParkNet were deployed in city taxicabs, the resulting mobile sensors would provide adequate coverage and be more cost-effective by an estimated factor of roughly 10-15 when compared to a sensor network with a dedicated sensor at every parking space, as is currently being tested in San Francisco.