Evaluation of Sensys Wireless Vehicle Detection System results from the first three months

Juan C. Medina, Ali Hajbabaie, R. F. Benekohal

Research output: Book/Report/Conference proceedingTechnical report


The Sensys Wireless Vehicle Detection System uses wireless magnetometers embedded in the pavement that communicate without wires to an access point connected to a standard detector rack. This research evaluated the detection performance of this system at two locations: a signalized intersection with three approaching lanes (at stop bar and advance zones), and a railroad grade crossing, as a potential backup option for crossing gates applications. Video images and the time of activation/deactivation of the sensors and loop detectors (placed at exactly the same location as the sensors), were collected after the system was installed by Sensys Networks, yielding some initial results (initial setup). Subsequently, Sensys Networks was provided with the initial results (false, missed, stuck-on, and dropped calls) and were allowed to readjust the system (modified setup) so that the best performance could be obtained. Thus, this report includes results from the initial and the modified setups at the two selected locations. At the railroad location, false calls due to adjacent vehicles ranged from 12.1% to 53.7%, missed calls were low (<1%), and stuck-on calls due to trains and cars were rare (but up to 30 min long). At the signalized intersection, false calls were more frequent at the stop bar (13.5% to 19.6%) than at the advance zones (0.7% to 2.4%). Missed calls were low at the stop bar (<0.5%), and ranged between 0.9% and 10% at the advance zones. Stuck-on calls and dropped calls were very rare and only found at stop bar zones.
Original languageEnglish (US)
Place of PublicationUrbana, Ill
PublisherIllinois Center for Transportation
StatePublished - 2009

Publication series

NameICT Series


Dive into the research topics of 'Evaluation of Sensys Wireless Vehicle Detection System results from the first three months'. Together they form a unique fingerprint.

Cite this