In-vehicle detection and warning devices

Several related concepts 'Driver Vigilance Monitoring', 'Drowsiness Detection Systems', 'Fatigue Monitoring Systems' refer to in-vehicle systems that monitor driver and/or vehicle behaviour. These systems monitor the performance of the driver, and provide alerts or stimulation if the driver seems to be impaired.

Driver and vehicle monitoring systems may monitor both driver and vehicle behaviour. Information can be gathered from driver input and control of the vehicles lateral position and speed, such as acceleration, steering wheel movement and lane position. Likewise, user behaviour such as eye movement, facial feature movement, brain waves (EEG) and steering wheel grip may all be monitored.

Estimations of the approximate reductions expected with lane driver monitoring systems in Germany (assuming 70% penetration of the passenger vehicle fleet) were reported by eSafety Forum [30]. It was expected that 50% of fatigue-related crashes would be affected, leading to a 35% reduction in these crashes. This would equate to a 2.9% reduction in all crashes.

Fatigue warning systems (FWS) have been proposed as specific countermeasures to reduce collisions associated with driver fatigue. These devices employ a variety of techniques for detecting driver drowsiness while operating a vehicle and signal a driver when critical drowsiness levels are reached. However, the detection of driver fatigue using valid, unobtrusive, and objective measures remains a significant challenge. Detection techniques may use lane departure, steering wheel activity, ocular or facial characteristics.

Several authors point out that fatigue warning systems may result in driver behavioural adaptation [98]. A possible negative effect of in-car warning systems may be that driver's use them to stay awake and drive for longer periods rather than stopping and have a nap; i.e. risk compensation by relying too much on the safety system.

This was confirmed by a study of Vincent, Noy & Laing [107]. They evaluated a fatigue warning system that measured ocular and face monitoring, vehicle speed, steering position and lane position. They found the users of the system did not take more or longer breaks, and did not show different fatigue levels to controls. Drivers generally ignored the FWS signals received. The physical aspect of the warning signals used in the study had no impact on driver fatigue levels. Voluntary rest stops, lasting on average 30 minutes, only had a minor impact on decreasing driver fatigue with short-lived effects. The authors concluded that voluntary breaks were ineffective in substantially counteracting the effects of fatigue associated with prolonged night time driving. Whereas normally rested drivers may successfully use breaks to prevent or postpone fatigue during a daytime drive [91], the use of breaks seems less successful in reducing fatigue resulting from prolonged nigh time driving and associated sleep loss.

In Europe, the project AWAKE has furthered our knowledge about driver vigilance systems. The EU project AWAKE (System for Effective Assessment of Driver Vigilance and Warning According to Traffic Risk Estimation) (http://www.awake-eu.org/) has developed guidelines for fatigue warning systems. A successful approach for on-road driver fatigue detection has to combine driver state and driver performance measures [111]. The AWAKE project has adopted this approach.

The AWAKE project aimed to demonstrate the technological feasibility of driver vigilance monitoring systems. To do this the project also looked at the non-technical issues that influence the use of such systems. The theoretically developed AWAKE system employs both driver state measures and traffic risk measures to arrive at a conclusion about the need for warning the driver and the type of warning called for. The driver state measures include eyelid movement, changes in steering grip and driver behaviour (including lane tracking), use of accelerator and brake, and steering position. These measures are input to a driver warning system that determines if and what information or warning messages need to be communicated to the driver. "Traffic risk estimation" data are used to re-assess driver's state and consequently, re-assess the conclusion about the type of warning needed. The risk of the traffic situation is estimated via a combination of data from digital navigation maps, anti-collision devices, driver gaze sensors and odometer readings. The project has produced several design guidelines for the assessment of driver vigilance and warning signals. These guidelines are fairly comprehensive. Although they do not address all questions, they are likely to influence future implementation of fatigue detection devices.

Despite progress made by the AWAKE project, there is still no golden standard and reference data for micro-sleep phenomenon. This hinders the success chances of further developing the tested systems. First, further research is needed to identify physiological data that are better able to distinguish between various states of sleepiness, inattention or stress). Currently, no single method exists that is commonly accepted to detect driver fatigue. Wright et al [113] have evaluated the sensitivity, intrusiveness, operational and market status of sleepiness detection devices. A subset of 15 devices was identified as being worthy of further evaluation.