Active driving assistance (ADA) systems are becoming increasingly popular and sometimes marketed under confusing names leading some consumers to overestimate system capabilities.
Article continues below
Currently available ADA systems are classified by SAE International®1 as a Level 2 partial driving automation feature, meaning that constant driver supervision is required, American Automobile Association, (AAA) reports.
To mitigate misuse of these systems, a driver monitoring component is integrated.
Within this study, driver monitoring systems are classified as either direct or indirect.
Direct systems integrate a driver-facing camera to detect driver distraction or disengagement.
In contrast, indirect systems only utilize steering wheel input for the detection of driver distraction or disengagement.
To evaluate the performance of driver monitoring systems with respect to system type, four popular vehicles equipped with an ADA system were evaluated by simulating driver disengagement (common behaviors such as texting, reading, watching videos, or general manipulation of a mobile device) in a real-world highway environment.
Direct driver monitoring systems were significantly more effective at mitigating driver disengagement than indirect driver monitoring systems in all lighting conditions.
On average, the percentage of time drivers were engaged was approximately five times greater for direct systems than indirect systems.
The findings of this study illustrate that direct driver monitoring systems are more effective than indirect systems at mitigating various types of driver disengagement.
Specifically, direct systems issued alerts faster and more persistently than indirect systems for both simulated driver disengagement modes and active circumvention attempts, on average.
Evaluated indirect systems were much less effective at mitigating ADA system misuse.
For fixed driver disengagement modes, indirect systems allowed over 50 seconds of simulated disengagement between alerts, on average.
For active circumvention attempts, indirect systems allowed an average of nearly 6 minutes of continuous simulated driver disengagement between alerts.
While evaluated direct systems exhibited superior performance, it is nonetheless concerning that researchers were able to continuously simulate disengagement over ten minutes without disablement of the ADA system.
It is acknowledged that researchers immediately discontinued simulated disengagement once an alert was provided, preventing an escalation of alerts.
All researchers accumulated dozens of alerts for direct systems within a ten-minute evaluation period for both disengagement modes evaluated.
AAA recommends that ADA systems become disabled for the remainder of the drive if too many initial alerts are provided within a given time period, independent of the requirement that drivers ignore a warning one or more times before disablement of the ADA system. ■