Naturalistic field evaluation of smartphone-based driver safety alerts in Gainesville, Florida used eye tracking and trajectory data.
Gainesville, Florida, United States
Evaluation of Intelligent School Zone Beacon and Vehicle-Cyclist Detection and Warning System
Summary Information
In a proactive effort to enhance safety in the vicinity of schools, Florida has included a requirement to install flashing beacons at reduced speed school zones. To address potential distracted drivers in school zones, several approaches have emerged, including a technological solution using Global Positioning System (GPS) location capability within smartphones to provide real-time alerts and warnings to drivers that do not reduce speed within a school zone. Florida DOT commissioned a study to evaluate one such solution, the TravelSafely app which alerts drivers if they exceed a speed threshold while in an active school zone. The technology relies on intelligent school zone beacon infrastructure, which provide the status of the school zone (active or not) to service providers. In addition, this app also provides a capability to provide drivers with alerts of bicyclists who also have the app active.
Methodology
Four school zones equipped with beacons were selected for inclusion in a 5.7-mile driving circuit, and 50 participants drove through the circuit twice in an instrumented vehicle with the researcher-provided smartphone with app. A staged bicyclist with the app (part of the research team) was also included to test the bicyclist component of the alert. Drivers were randomly assigned to one of three conditions:
- Stealth/OFF condition (no alerts)
- Audio ON (driver received audio alerts)
- Audio/Visual ON (driver received both audio and visual alerts)
For drivers experiencing the ON condition, alerts were triggered if the driver exceeded 20 mph in the school zone, which all had 15 mph limits. A mix of manual and automated methods was used to analyze driving behavior and locations of eye glances. GPS trajectory data was used to assess occurrences of speeds in excess of 20 mph within the school zone limit area.
Findings
In all four tested school zone locations, the probability of drivers exceeding the 20 mph threshold was lower when the app alerts were active. The following table shows the results for each zone and condition:
Table: Probability of speeds exceeding 20 mph (in 15 mph school zone)
|
Condition |
Zone 1 |
Zone 2 |
Zone 3 |
Zone 4 |
|
Stealth/OFF |
36% |
6.06% |
36% |
14% |
|
Audio ON |
23% |
4.6% |
13.4% |
11% |
|
Audio/Visual ON |
23% |
1.1% |
1% |
6.4% |
The analysis of eye glance behavior showed increase visual scanning behavior for drivers with the app active. Together with speed reductions, the behaviors could translate into improved situational awareness and safety in school zones.
For the alert component based on the bicyclist with app active, researchers found an increase in probability of seeing the bicyclist when not expected (in the first trip through the loop). In the second loop through the circuit (bicyclist expected), the probability of the driver seeing the bicyclist was not significantly different as the bicyclist was likely to be noticed without the app.
