In-Vehicle Secondary Task Study Based On Human-Machine Interactive Simulation

Driving safety is a global public safety issue, towards which studies should be focused on driver's characteristics, attitude, behavior and capacity. With the boom of In-Vehicle Information System (IVIS), in-vehicle secondary task has become a major challenge for driving safety by competing on visual and cognitive resources with primary driving tasks. Driving simulation was adopted by many researchers in this field as an ideal tool because of its advantages in easy scenario design, repeatable testing, easy data acquisition, and ensuring testing safety. However, the validity of simulation method lacks sounding supportive studies; meanwhile, the effectiveness of simulation method in engineering application for safety promotion was also a key area to explore. These two problems were studied in this research based on selected typical visual and cognitive secondary tasks.In the visual secondary task study, IVIS operations were both tested on simulator and real driving on road to assess the validity of both the simulation method and the resource requirement of three IVIS input devices. The results show that both relative and absolute validity were satisfied for the measurement of driver's visual attention and IVIS task performance. The two measured indices also show great sensitivity on IVIS assessment. Contrarily, driving performance of longitudinal and lateral vehicle control only established low validity and poor sensitivity. Among the three devices, keypad presented the best usability and safeness, scrolling wheel was the most demanding, and touch screen needs most visual resources. Additionally, stationary test of IVIS task performance has lower precision than simulation, but task duration was suggested to be used as testing index with satisfactory precision when simulation or on-road test are not possible.In the cognitive secondary task study, a participative feedback training method towards driver's cell phone use attitude was developed based on simulation. The training was proved effective in changing driver's self-calibration on driving degradation caused by distraction and changing driver's attitude towards cell phone use for both short-term and long-term. Compared to basic driving performance, performance of recognition and detection tasks showed more degradation. Accordingly, the latter tasks are suggested when designing similar training programs. The results from the visual task study could provide a sounding validity support for other researchers in driving safety study using simulation method and provide helpful reference data and method on IVIS design and evaluation for practitioners. Meanwhile, the cognitive task study greatly encouraged positive interventions on driver's attitude and behavior through simulation-based driving training program in future.