A field trial has been carried out using a set of automatic
recording systems, Driving Monitoring Recorder (DMR) with GPS and
Event Eye Camera (EEC). They were installed on total numbers of 105
vehicles in four fleets of taxis and trucks in Tokyo area for 4
years in order to assess the implications in driving characteristic
and traffic conditions.
DMR can record the data such as running speed versus time
continually, and frequencies of emergency behavior are counted and
recorded on drive whenever hard braking is exceeded by 3.75m/s 2,
and rapid starting and sharp acceleration is exceeded by 3.5 m/s 2.
Also, EEC can record the driver views in pre- and post scene during
10 seconds by event detecting trigger. The 80 subjects are taxi
drivers including 10 females and 30 truck drivers. The drivers were
monitored for some months continuously. In addition, the drivers
were examined by a driving aptitude test.
The analysis result indicates that individual drivers have a
specific driving performance on drives, and it suggest that drivers
will be able to specified their driver characteristics by DMR data.
For simplicity, we defined the "'Index of bad
behavior'' score as simply adding points of numbers of the
three emergency actions per 100 km running. The drivers are
classified into 4 groups, ranked A, B, C and D levels according the
index score, which ranging from careless driver to model of safety
driver. The ranks are compared with the results of driving aptitude
test. To assess driver behavior, EEC images were examined when,
where and what condition emergency behavior occurred. As results of
EEC image analysis, some drivers with high bad index scores had
driven at the head mostly, and frequent rapid lane changing and
sharp turn actions.
In conclusion, a set of the automatic recording systems can
offer useful data to study driver characteristics. In addition, the
important aspect of this study is persuasively, most drivers tend
to accept these data without complaining to improve their safety
drive because of digital data and image data.