Driver Demand: Eye Glance Measures

2016-01-1421

04/05/2016

Event
SAE 2016 World Congress and Exhibition
Authors Abstract
Content
This study investigated driver glances while engaging in infotainment tasks in a stationary vehicle while surrogate driving: watching a driving video recorded from a driver’s viewpoint and projected on a large screen, performing a lane-tracking task, and performing the Tactile Detection Response Task (TDRT) to measure attentional effects of secondary tasks on event detection and response. Twenty-four participants were seated in a 2014 Toyota Corolla production vehicle with the navigation system option. They performed the lane-tracking task using the vehicle’s steering wheel, fitted with a laser pointer to indicate wheel movement on the driving video. Participants simultaneously performed the TDRT and a variety of infotainment tasks, including Manual and Mixed-Mode versions of Destination Entry and Cancel, Contact Dialing, Radio Tuning, Radio Preset selection, and other Manual tasks. Participants also completed the 0-and 1-Back pure auditory-vocal tasks. Glances were recorded using an eye-tracker, and validated by manual inspection. Glances were classified as on-road (i.e., looking through the windshield) or off-road (i.e., to locations other than through the windshield). Three off-road glance metrics were tabulated and scored using the NHTSA Guidelines methods: Mean Single Glance Duration (MSGD), Total Eyes-Off-Road Time (TEORT), and Long Glance Proportion (LGP). Comparisons were made for these metric values between the task conditions and a 30-s Baseline condition with no task. Mixed-Mode tasks did not have a statistically significant longer MSGD or TEORT, or higher LGP, than Baseline (except for Mixed-Mode Destination Entry), whereas all the Manual tasks did. Mixed-Mode tasks improved compliance with the NHTSA Guidelines.
Meta TagsDetails
DOI
https://doi.org/10.4271/2016-01-1421
Pages
9
Citation
Seaman, S., Hsieh, L., and Young, R., "Driver Demand: Eye Glance Measures," SAE Technical Paper 2016-01-1421, 2016, https://doi.org/10.4271/2016-01-1421.
Additional Details
Publisher
Published
Apr 5, 2016
Product Code
2016-01-1421
Content Type
Technical Paper
Language
English