Finding manoeuvre motifs in vehicle telematics

Area of Science:

• Road safety research
• Data science applications in transportation

Background:

• Driving behavior significantly influences road safety.
• Analyzing driving maneuvers provides key insights into driver actions.

Purpose of the Study:

• To investigate a novel method for identifying driving maneuvers using vehicle telematics data.
• To apply motif detection in time-series analysis for maneuver recognition.

Main Methods:

• Implementation of a modified Extended Motif Discovery (EMD) algorithm.
• Application of the EMD algorithm to the UAH-DriveSet naturalistic driving dataset.
• Systematic exploration and analysis of extracted time-series motifs.

Main Results:

• The EMD algorithm effectively identified simple driving maneuvers like accelerations, braking, and curves.
• Complex maneuvers, including lane changes and overtaking, were also successfully extracted.
• Validation of motif discovery as a viable approach for analyzing driving behavior.

Conclusions:

• Motif detection is a promising technique for understanding driving maneuvers from telematics data.
• The EMD algorithm demonstrates capability in recognizing a range of driving actions.
• Further research into motif discovery for road safety analysis is warranted.