Driving Behavior Research for Intelligent Collision Avoidance Technology

Work Summary

  • Examined the effects of situational urgency on drivers’ collision avoidance behaviors using Tongji University’s eight-degree-of-freedom driving simulator.
  • Developed a kinematic-based forward collision warning (FCW) algorithm that is compatible with drivers’ risk perceptions and behavioral responses.
  • Implemented the proposed FCW algorithm in Tongji University driving simulator, and evaluated the system’s performance, warning timing, and safety benefits.


Driving Behavior Research for Intelligent Collision Avoidance Technology is a driving simulator study jointly conducted by Tongji University and China First Automobile Work (FAW) Corporation (one of the largest automobile companies in China). The main purpose of this project is to improve the understanding of drivers’ collision avoidance behavior under different rear-end scenarios and to develop an effective forward collision warning (FCW) strategy. Driving simulators are ideal for performing these kinds of studies because of their ability to systematically vary perceived urgency while capturing quantitative data on relevant aspects of driver and system performance.

In this project, the Tongji University Driving Simulator was used to generate different urgency levels by varying headway and LV deceleration while capturing data on perception response times (PRT), throttle release response times, throttle to brake transition times, brake delays, maximum brake pedal pressures and peak decelerations. The relationships uncovered between situational urgency and drivers’ collision avoidance behavior measures can provide information that can be used to develop improved FCW systems.


The Tongji University driving simulator, currently the most advanced in China, incorporates a fully instrumented Renault Megane III vehicle cab in a dome mounted on an 8 degree-of-freedom motion system with an X-Y range of 20 × 5 meters. An immersive 5 projector system provides a front image view of 250° × 40° at 1000 × 1050 resolution refreshed at 60 Hz. LCD monitors provide rear views at the central and side mirror positions. SCANeRTM studio software (OKTAL) presented the simulated roadway and controlled a force feedback system that acquired data from the steering wheel, pedals and gear shift lever. The transmission of the Renault Megane III vehicle was automatic, and the braking system was a non-ABS. The overall performance of this driving simulator was validated using three tests: simulator sickness, stop distance, and traffic sign size. Test results showed that the driving simulator satisfied the three criteria (i.e. at least 75% of participants show no simulator sickness, stop the car within 2 meters of a designated stop line and judge the realism of the traffic sign size) for validation.

Tongji University driving simulator


Six females and 23 males, (ages 23–54, M = 33.2, SD = 8.3), who possessed valid driver’s licenses and had at least one year and 10,000 kilometers of driving experience recruited from the population of drivers in Shanghai served as participants. Participants initially drive on the inner lane of a two-lane freeway under good weather daytime conditions with light traffic, and were asked to accelerate to the target speed (120 km/h) at the beginning of the scenario. To minimize distractions, traffic was not present in the same direction of the SV, although for realism, light traffic was presented in the opposite direction. After about 2 minutes, a white lead vehicle (LV) moved in front of the SV. The LV was programmed to operate at a constant speed of 120 km/h, and participants were again asked to follow the lead vehicle at distance of 60 m to 80 m. The LV was programmed to make 6 unpredicted full stops with brake lights on, at prearranged initial headway settings of 1.5 sec and 2.5 sec, and at varying intervals.

Video monitor displays (a) and experiment scenario (b)

To reduce the predictability of LV stops, there were two cases during the test phase where LV slowed down but with small deceleration rates of 0.02 g to 0.1 g. When the LV was triggered to stop, if the control program determined the SV was not within the specified headway range, a “Speed Up” message was displayed on the screen until the SV reached the targeted headway. To prevent drivers from anticipating collision situations in association with “Speed Up” messages, instances were included in the experiment in which the “Speed Up” message was displayed but without a subsequent sudden LV brake. A minimum period of 5 seconds was then introduced during which it was confirmed that the participants were following the LV steadily. Once confirmed, the LV would come to a stop at the programmed deceleration rate. It should be noted that all the programmed events occurred on flat straight roads, thus eliminating the effects of horizontal curves and longitudinal slopes on drivers’ braking and steering. Test phases were completed after 6 full stops were made, and required about 30 minutes. A post-simulation survey of participants conducted showed that more than 60% of drivers said the vehicle dynamics, motion systems, and visual and audio systems of the driving simulator had a high level of realism. Throughout the experiment, participants were visually monitored using four video cameras. The following figure shows the sequence of timed events and example curves for vehicle speed, acceleration, throttle, steering wheel angle, and braking pedal pressure as they change during a collision avoidance episode.

A typical collision avoidance event sequence


. Dimension Reduction and Multivariate Analysis of Variance for Drivers’ Forward Collision Avoidance Behavior Characteristics. Journal of Tongji University, Vol. 44, No. 12, pp. 1858-1866, 2016.

PDF Project

. Impacts of Situational Urgency on Drivers’ Collision Avoidance Behaviors. Journal of Tongji University, Vol. 44, No. 6, pp. 876-883, 2016.

PDF Project

. Drivers’ Rear End Collision Avoidance Behaviors under Different Levels of Situational Urgency. Transportation Research Part C: Emerging Technologies, Vol. 71, pp. 419-433, 2016.

PDF Project Source Document

. Development of a kinematic-based forward collision warning algorithm using an advanced driving simulator. IEEE Transactions on Intelligent Transportation Systems, Vol. 17, No. 9, pp. 2583-2591, 2016.

PDF Project Source Document