ABSTRACT
OBJECTIVE: Varying driver distraction algorithms were developed using vehicle kinematics and driver gaze data obtained from a camera-based driver monitoring system (DMS). BACKGROUND: Distracted driving characteristics can be difficult to accurately detect due to wide variation in driver behavior across driving environments. The growing availability of information about drivers and their involvement in the driving task increases the opportunity for accurately recognizing attention state. METHOD: A baseline for driver distraction levels was developed using a video feed of 24 separate drivers in varying naturalistic driving conditions. This initial assessment was used to develop four buffer-based algorithms that aimed to determine a driver's real-time attentiveness, via a variety of metrics and combinations thereof. RESULTS: Of those tested, the optimal algorithm included ungrouped glance locations and speed. Notably, as an algorithm's performance of detecting very distracted drivers improved, its accuracy for correctly identifying attentive drivers decreased. CONCLUSION: At a minimum, drivers' gaze position and vehicle speed should be included when designing driver distraction algorithms to delineate between glance patterns observed at high and low speeds. Distraction algorithms should be designed with an understanding of their limitations, including instances in which they may fail to detect distracted drivers, or falsely notify attentive drivers. APPLICATION: This research adds to the body of knowledge related to driver distraction and contributes to available methods to potentially address and reduce occurrences. Machine learning algorithms can build on the data elements discussed to increase distraction detection accuracy using robust artificial intelligence.
ABSTRACT
Detection performance as a function of distance was measured for 16 subjects who pressed a button upon aurally detecting the approach of an electric vehicle. The vehicle was equipped with loudspeakers that broadcast one of four additive warning sounds. Other test conditions included two vehicle approach speeds [10 and 20 km/h (kph)] and two background noise conditions (55 and 60 dBA). All of the test warning sounds were designed to be compliant with FMVSS 141 proposed regulations in regard to the overall sound pressure levels around the vehicle and in 1/3 octave band levels. Previous work has provided detection results as average vehicle detection distance. This work provides the results as probability of detection (Pd) as a function of distance. The curves provide insight into the false alarm rate when the vehicle is far away from the listeners as well and the Pd at the mean detection distance. Results suggest that, although the test sounds provide an average detection distance that exceeds the National Highway Traffic Safety Administration minimum at the two test speeds, Pd is not always 100% at those distances, particularly at the 10 kph. At the higher speed of 20 kph, the tire-road interaction noise becomes dominant, and the detection range is greatly extended.
