Driver Workload Metrics: Experimental Data from Laboratory, Highway, and Test Track

A study by Linda Angell presents driver performance data collected from 57 drivers in a laboratory, 108 drivers on an interstate highway, and 69 drivers on a test track. The research examined 22 in-vehicle tasks plus baseline driving under various conditions to develop metrics and a toolkit for assessing visual, manual, and cognitive workload. Analysis focused on selecting repeatable and predictive metrics from lateral/longitudinal control, object detection, and eyeglance behavior.

Use Cases

• Developing predictive models of driver distraction based on the described visual, manual, and cognitive workload dimensions.
• Benchmarking in-vehicle system safety using the reported metrics for lateral/longitudinal control and object detection.
• Training machine learning classifiers to discriminate between high and low workload tasks using the experimental performance profiles.
• Validating simulation tools against the real-world driving data collected on the highway and test track.

Strengths

• Data was collected from three distinct venues (laboratory, highway, test track) with 57, 108, and 69 drivers respectively.
• Analysis focused on selecting metrics that were described as repeatable, predictive, and capable of discriminating workload levels.
• The study examined 22 specific in-vehicle tasks plus baseline driving under a variety of experimental conditions.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

Linda Angell

Collection Method

Driver performance data were collected in laboratory, on an interstate highway, and on a test track under controlled experimental conditions.

Time Range

null

Freshness

Last updated date is unknown.

Geography

null