Driver Identification Leveraging Single-turn Behaviors via Mobile Devices

Yan Wang, Tianming Zhao, Fatemeh Tahmasbi, Jerry Cheng, Yingying Chen, Jiadi Yu

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Scopus citations


Drivers' identities are essential information that can facilitate a broad range of applications. For example, by understanding who is driving the vehicle when an accident happens, insurance companies could determine the liability and payment in a car accident claim case with high confidence. Another example, pick-up service companies could track the identities of their drivers to ensure that authorized drivers are driving esteemed clients to their destinations. While there are existing studies that can utilize video cameras and dedicated sensors to identify drivers, they either have privacy issues or require additional hardware, which is not practical enough for daily uses. In this paper, we devise a low-cost driver identification system, which can determine drivers' identities by using sensors readily available in wearable devices. Our system captures the unique driving behaviors during pervasive but momentary driving events (i.e., turning at intersections) with motion sensors, which are widely integrated into commodity wearable devices (e.g., smartphones and activity trackers). Toward this end, we extensively analyze people's driving behaviors and identify the critical turning events that capture people's unique behavioral patterns for driver identification. We design a fine-grained turning segmentation method that divides sensor data into critical turning stages (i.e., before, during, and after-turn stages), which provide multiple dimensions of turning behavioral metrics facilitating driver identification. The system extracts unique turning behavior features from time and frequency domains to enable driver identification based on drivers' turning behaviors at different types of turns. Extensive experiments are conducted with 12 drivers and various types of turns in real-road conditions. The results demonstrate that our system can identify drivers with high accuracy and low falsepositive rate based on one single turning event.

Original languageEnglish (US)
Title of host publicationICCCN 2020 - 29th International Conference on Computer Communications and Networks
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781728166070
StatePublished - Aug 2020
Event29th International Conference on Computer Communications and Networks, ICCCN 2020 - Honolulu, United States
Duration: Aug 3 2020Aug 6 2020

Publication series

NameProceedings - International Conference on Computer Communications and Networks, ICCCN
ISSN (Print)1095-2055


Conference29th International Conference on Computer Communications and Networks, ICCCN 2020
Country/TerritoryUnited States

All Science Journal Classification (ASJC) codes

  • Computer Networks and Communications
  • Hardware and Architecture
  • Software


  • Driver Identification
  • Smartphone


Dive into the research topics of 'Driver Identification Leveraging Single-turn Behaviors via Mobile Devices'. Together they form a unique fingerprint.

Cite this