Road accidents statistics clearly pose achallenge on three main aspects: Human Error, Vehicle failures and Roadconditions. Human errors being most influential factor for cause of accidents,serious attention needs to be paid on detection, analysis and monitoring ofdriver behaviour. Understanding driver’s behaviour is the key factor whichcontributes towards the road safety. Moreover if the driver’s behaviour isrecorded and analyzed, it could have influential positive impact on the system.
Hence the methods for detection and monitoring of behaviour exhibited by driverare of significant interest. By doing so, compliance to driving regulationcould be achieved and this may help achieve the goal of road and driver safety. Driver has to dynamically interact to theroad surface and traffic conditions to control the vehicle. It is very naturalthat different driver’s will respond differently to the similar drivingconditions basically through exhibiting acceleration, breaking and steering.Vehicle Driver’s should be able to identify the vehicle dynamics in terms ofposition, velocity, acceleration, orientation and direction of the vehicle and alsokeep a watch whether change in the relationship between these factors isleading to an risky condition either for occupants of the vehicle or other roadusers.
The most stable ride is thesafest ride. In most of the cases it becomes unrealistic to set a same hardlimit for all vehicles or for all drivers. As every driver shows a behaviourwhich is unique, it becomes necessary to change the limits based on the type ofvehicle and the limit can be specific to category of them. By collecting andanalyzing the vehicle manoeuvring data, driving behaviours could be classifiedas aggressive or normal behaviour. To ensure overall stability in vehiclemotion, an indicative measure is of high significance.
For fleet management applications identifyingabnormal driving manoeuvres is definitely an important research focus due toits significant impact on fuel consumption and road safety. The in-vehiclesensing and Internet-of-Vehicles (IoV) technologies put together are capable ofcollecting abundant IMU data viz. longitudinal accelerations and lateralacceleration, driving data such as speed, steer angle and engine parameters,from a large number of vehicles.
Such data are categorized as large volume, multidomain, multi-frequency, and multi-source, which mainly reflect the vehiclestatus and thereby are extensively used to assess driving behaviours. (Mingminget al., 2017) Some insurance companies provide extendedwarranty on range of parts in vehicle and it becomes necessary for them to knowwhether these parts were used properly as advised. The advantage of detectingthe abnormal driving behaviours for insurance companies is providing a new’pay-as-you-drive’ service to clients by collecting dynamic data and judgingtheir driving manoeuvres.
Collected data is then analyzed and thus fleet-operatingcompanies can regulate their drivers to act more wisely while riding car,lowering the accidental risk and fuel consumption. Continuous tracking ofbehaviour of driver will involve feedback from fleet manager and hence thiswill assist driver to further increase the fuel efficiency and also be a safedriver. This will also assure that comfort level of customers will drasticallyincrease. All the above are possible through Cartravelling data recorder or Car Black Box which plays an important role inpreventing fatigue driving, over speed and motoring offences, restricting thedriver’s malpractice, analyzing the accident, enforcing traffic management andtransportation, as well as ensuring driving safety of the car.