Estimating pedestrian traffic with Bluetooth sensor technology

Avital Angel, Achituv Cohen, Sagi Dalyot, Pnina Plaut

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

The increasing availability of ubiquitous sensor data on the built environment holds great potential for a new generation of travel and mobility research. Bluetooth technology, for example, is already vastly used in vehicular transportation management solutions and services. Current studies discuss the potential of this emerging technology for pedestrian mobility research, but it has yet to be examined in a large urban setting. One of the main problems is detecting pedestrians from Bluetooth records since their behavior and movement patterns share similarities with other urban transportation modes. This study aims to accurately detect pedestrians using a network of 65 Bluetooth detectors located in Tel-Aviv, Israel, which record on average over 60,000 unique daily Bluetooth Media-Access-Control addresses. We propose a detection methodology that includes system calibration, effective travel time calculation, and classification by velocity that takes into consideration the probability of vehicular traffic jams. An evaluation of the proposed methodology presents a promising pedestrian detection accuracy rate of 89%. We showcase the results of pedestrian traffic analysis, together with a discussion on the data analysis challenges and limitations. To the best of our knowledge, this work is the first to analyze pedestrian records detection from a Bluetooth network employed in a dynamic urban environment setting.

Original languageEnglish
JournalGeo-Spatial Information Science
DOIs
StateAccepted/In press - 2023
Externally publishedYes

Keywords

  • Bluetooth technology
  • pedestrian detection
  • pedestrian mobility
  • ubiquitous sensor network
  • walking

Fingerprint

Dive into the research topics of 'Estimating pedestrian traffic with Bluetooth sensor technology'. Together they form a unique fingerprint.

Cite this