Using Technology to Eliminate Drowsy Driving

Abstract

Operator fatigue is an issue in all 24-hour operations; operator fatigue is a leading cause of vehicle crashes. The oil and gas industry is a 24-hour operation that relies extensively on land transportation. It comes as no surprise then that drowsy driving has been recognized as a cause of many land transportation-related fatalities. Dealing with drowsy driving is a difficult task. Training is a powerful tool; however, drivers trained to recognize and manage fatigue frequently fail to effectively apply this training at critical times because sound judgment is one of the first things affected by fatigue. As part of an ongoing commitment to crash-free driving, a major oilfield service company turned to a leading fatigue research and technology company to assist in eliminating drowsy driving. This paper reports on the experiences of a field trial with a new fatigue warning device for drivers and operators. The device utilizes a camera system installed in the dashboard of the vehicle that monitors the eyes of the driver. In contrast to previously proposed devices, this system does not utilize infra-red (IR) reflections, and as a result, is able to operate during daytime and when eyeglasses are worn. Deployed in four heavy trucks and one light truck, the systems have successfully recorded sequences of the eyes of drivers during potential fatigue events. This paper also reports on staff acceptance of such fatigue management systems. Privacy concerns are an issue when camera-bsaed systems are used to monitor drivers. Introduction Fatigue and the resulting drowsiness is a key risk factor in land transportation operations, and for industries like the oil and gas industry in particular, where operations in many cases continue around the clock. While truck safety has increased through numerous safety systems and improvements of their mechanical design, key risk factors related to the driver have proved to be hard to address so far. Research in recent years suggests that fatigue and distraction are major factors in accident causation in transportation, and for heavy vehicles in particular, fatigue is a big contributor. A study [1] conducted by the National Transportation Safety Board concluded that 30% to 40% of all heavy truck crashes may be attributed to driver drowsiness. Another study [2], which involved extensive instrumentation of 100 passenger vehicles for one year, concluded that 78% of the 82 accidents recorded by the vehicles were attributed to driver inattention to the road ahead. In their work, inattention is defined as secondary task engagement, driving related inattention to the roadway, drowsiness, or non-specific eye glances, which encompasses all possible causes for inattention to the road ahead. In addition to the high proportion of fatigue-related heavy vehicle crashes, accidents as a result of fatigue are much more severe than accidents on average. The plot in Figure 1 shows the impact of fatigue on accident severity. The left column represents all vehicle accidents grouped according to the severity of the injury of the most severely injured person in the accident (Maximum Accident Injury Scale). About 90% of all accidents result in medium or lesser injuries, while less than 1% of all accidents are fatal. The right column represents only fatigue-related accidents, and only 65% of these result in medium or lesser injuries, while 7% result in fatalities. The reason for this increase in injury severity is that fatigue-related crashes often involve no or considerably delayed braking, and therefore, a higher average impact velocity. While this statistic is based on all road transport accidents, a similar trend can be assumed for heavy truck accidents.

Driver inattention monitoring system for intelligent vehicles: A review

“…As the image processing technique developed, some commercial products measuring driver's states are already available on the market, such as smartEye AntiSleep [39,40] and SeeingMachines DSS [41,42]. Both of them use two IR illuminators to enhance their robust to light condition, and employ only one camera giving 3D information.…”

Open-eye detection using iris–sclera pattern analysis for driver drowsiness detection

“…These approaches effectively measure physical changes of the driver such as sagging posture, leaning of the driver's head and the open/closed states of the eyes. Different kinds of cameras and analysis algorithms have been reported in the literature for this approach: methods based on visible spectrum camera [16,17]; methods based on infra-red (IR) camera [7,[18][19][20] and methods based on stereo camera.…”

Fusion of Optimized Indicators from Advanced Driver Assistance Systems (ADAS) for Driver Drowsiness Detection

“…Additionally, they do not work properly with users wearing glasses and need high computational requirements. In the state-of-the-art, different kinds of cameras and processing algorithms have been proposed: visible camera-based methods [ 13 , 30 – 32 ]; near-infrared (NIR) camera-based methods [ 33 – 35 ]; and methods based on stereo cameras [ 36 – 38 ]. Some of them are commercial products.…”