CYGNIFY ACTIVE DRIVEGUARD
Cygnify Active Driveguard is our advanced driver monitoring and protection product, designed to detect, predict, and counter driver distraction and fatigue. It uses small, off-the-self cameras and computer hardware which can be installed in any vehicle. These are combined with proprietary deep learning-based AI algorithms running real-time on that hardware.
Inward-looking cameras monitor driver facial expression, gaze, and eye closure and opening patterns, as well as (optionally) in-cabin secondary task engagement based on detection of activities such as cell phone use. Simultaneously, outward-looking cameras monitor driving behavior including lane departures and unsafe interactions with other road users. Because distraction and especially fatigue are expressed very differently by different drivers, the underlying AI models are automatically personalized in the early stages of the driver's use of the system; which is another key contributor to the system's high accuracy.
Cygnify collaborates with leading academic institutes doing research on fatigue and distraction VTI and SWOV, and has unique access to several very large datasets with scientifically validated fatigue and distraction ground truth data. For distraction, this includes access to Europe's largest naturalistic driving database, UDRIVE. UDRIVE contains tens of thousands of hours of driving data (including video data) by both private car and professional truck drivers, and many instances of real-world distraction. Similarly, for fatigue we have access to unique data of fatigued and non-fatigued drivers driving on real roads (i.e. not simulator data, as most research has been limited to), with solid and frequent ground truth fatigue assessments. These datasets have allowed optimization of Cygnify Active Driveguard's detection accuracy to unparalleled accuracy; e.g., real-time detection of fatigue with better than 90% accuracy. Also uniquely, this high accuracy has been comfirmed in independent evaluation by our academic partners.
The modular set-up of Cygnify Active Driveguard enables deployment in various forms, allowing both real-time warnings and back-end management dashboard information, and allowing diverse customized software and hardware configurations delivered as full solutions by Cygnify.
Demonstration of Cygnify Active Driveguard's distraction detection algorithms. The system extracts information on the driver's eye gaze (eyes off road vs. on road), detected in-cabin objects such as cell phones, and manipulation of such objects (activity recognition). All information is combined to detect distraction, categorize the type of distraction, and assess its severity.
Demonstration of Cygnify Active Driveguard's fatigue detection algorithms. The system extracts information on the driver's eye closure and eye gaze patterns, facial expressiveness patterns including yawning, head nodding, etc., as well as lane center deviation patterns. All information is combined to detect fatigue and assess its severity.