100-Car Naturalistic Driving Study: Phase II Results of a Large-Scale Field Experiment

Phase II of a three-phase effort collected approximately 2,000,000 vehicle miles and almost 43,000 hours of naturalistic driving data over 12 to 14 months per vehicle. The study, sponsored by the National Highway Traffic Safety Administration and the Virginia Department of Transportation, involved 241 primary and secondary drivers using instrumented vehicles, 78 of which were privately owned. The resulting database contains video and sensor data on extreme driving behaviors, crashes, near-crashes, and incidents, classified into an epidemiological-style event database.

Use Cases

• Analyze driver reaction times and vehicle kinematics based on recorded parameters like speed and time-to-collision.
• Study the role of inattention and secondary tasks in crash and near-crash events based on video and event classification data.
• Model risk factors for rear-end and lane change events based on pre-event maneuvers and precipitating factors documented in the event database.
• Develop behavioral profiles for aggressive driving or drowsiness based on the database's inclusion of many extreme cases of driving performance.

Strengths

• Large-scale data collection with approximately 2,000,000 vehicle miles and almost 43,000 hours of data.
• Rich multimodal data from a highly capable instrumentation system including 5 channels of video and many vehicle state and kinematic sensors.
• Naturalistic study design where drivers were given no special instructions and data collection was unobtrusive, likely increasing ecological validity.
• Database contains many documented extreme cases of driving behavior, including severe drowsiness, impairment, and traffic violations.

Limitations

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

Provenance

Source

National Highway Traffic Safety Administration and Virginia Department of Transportation, authored by Thomas A. Dingus.

Collection Method

Field experiment using instrumented vehicles, with data collected unobtrusively from primary and secondary drivers over 12-14 months.

Time Range

Data collection lasted 12 to 14 months for each vehicle; specific years are not provided.

Geography

Likely focused on Virginia, USA, given the involvement of the Virginia Department of Transportation, but not explicitly stated.