Around 10-20% of all road accidents seem to be related to driver hypovigilance (fatigue or prolonged inattention). AWAKE project aims to increase traffic safety by reducing the number and the consequences of traffic accidents caused by driver's hypovigilance, by developing a multi-sensor system, including the most promising detection sensors developed so far, such as an eyelid camera, a steering grip sensor, a lane tracker and other sensors monitoring vehicle-related parameters. Their signals will be integrated, through intelligent algorithms, allowing strong system personalization to the specific traffic conditions as well to the driving pattern and physical characteristics of the individual driver. The system will be installed at three demonstrators (city car, luxury car and truck) and will be optimised through 18 pilots in 6 European countries, leading to an efficient and user-friendly driver warning system and to design guidelines and policy recommendations for its implementation.
- To significantly reduce accidents related to driver's hypovigilance and enhance user acceptance of driver monitoring systems;
- To develop a module, that will detect and diagnose in real-time driver hypovigilance;
- To develop an optimum, modular, on-time driver warning system, to alert the driver in case of drowsiness, using acoustic, visual and haptic means with different warning levels according to driver's vigilance status and the estimated risk level of the situation;
- To develop a module, to estimate the actual traffic risk and adapt a accordingly the provided warning to the driver;
- To integrate all these modules in a middle, upper class passenger car and a heavy vehicle demonstrator and profoundly test them in 16 pilots all over Europe.
The user needs regarding driver monitoring will be carefully analysed, with emphasis on priority driver cohorts, like young drivers, shift workers, professional drivers and people suffering from sleep disorders. Also the relevant legal and insurance issues will be deeply studied. Based on the above, the AWAKE system architecture and functional specifications will be drafted, whereas technical and other risk will be clearly monitored by risk analysis. An Hypovigilance Diagnosis Module (HDM) for highway scenarios will be developed, based on an existing prototype, by data fusion of the most relevant sensors data, such as eyelid movement, steering grip and lane tracker as well as other vehicle-related parameters. Data fusion will be primarily stochastic-based, but in parallel a deterministic HDM version will be realised, to act as filter to the probabilistic one but also to constitute its fall-back position. A Traffic Risk Estimation (TRE) module will be developed, based on existing driver assistance systems and sensors, by correlating vehicle sensor data on imminent risk with a map database of critical landmarks and driver's risk awareness through a driver's gaze detection sensor. Driver warning will be based on a modular approach, defining different warning strategies and elements, depending on the level of measured driver's hypovigilance, the estimated traffic risk, driver's type and application type and using a combination of visual, audio and kinesthetic elements (e.g. sound-emulated rumble strips).
The different modules will be integrated in three different prototypes: a city car, a luxury car and a heavy vehicle. All subsystems, sensors and the final system will be evaluated and optimised through 16 pilots all over Europe. The project work will also result in design guidelines for the industry, policy recommendations for the authorities, dissemination schisms as well as an interactive, s/w-based alertness education package and affordable products in the market for the general public.
A modular system (AWAKE), that can be installed in city and luxury passenger cars and heavy vehicles, to:- predict driver's hypovigilance in all highway scenarios with recognition rate over 90% and false alarm rates below 1% - successfully and safely warn the driver about hypovigilance state.
Funding SchemeCSC - Cost-sharing contracts
447 83 Vaargaarda
75794 Paris Cedex 16
75794 Paris Cedex 16
10043 Orbassano (To)
581 95 Linkoeping
2628 BL Delft