Personalization of parking styles and steering assistance for reducing discomfort with parking assist systems

Kenta Maeda･Naoyuki Tashiro･Swarn Singh Rathour･Shinji Seto (Hitachi)･Daisuke Tsuga･Hiroki Sato･Miki Koso･Atsushi Yokoyama (Hitachi Astemo)

Parking is one of the driving tasks that many drivers feel difficult. However, the parking assist system also has the problem of a low utilization due to anxiety and discomfort caused by difference from individual driver’s driving style. In this presentation, we consider a method of reducing anxiety by assisting with the driver’s manual parking operation based on haptic feedback, and reducing discomfort by using a target trajectory that is adapted to the individual parking styles of the driver.

Development of Trajectory Prediction Methods for Oncoming Vehicles and Path Planning Techniques for Narrow Road Passing by Autonomous Driving

Ryo Inaba･Masato Imai･Shuntaro Tsuchiya･DANIEL GABRIEL (Hitachi)･Hidehiro Toyoda･Satoshi Ito･Yuichi Komoriya･Ryo Sakurai (Hitachi Astemo, Ltd.)

We are developing autonomous driving technology for highways, general roads, and residential roads. On residential roads, cooperative actions with other vehicles, such as merging and passing, are necessary without wireless communication. In this presentation, we propose a machine learning-based trajectory prediction method for oncoming vehicles and a path planning method including reverse evasion. We report the evaluation results of the proposed method at the road environment modelled on residential streets.

Trajectory Planning for Lane Change Feasibility Decision Using Chance-Constrained Model Predictive Control

Yuichi Okura･Kenta Tominaga･Hiroaki Kitano (Mitsubishi Electric)

The objective of this study is to improve the decision performance for lane change feasibility by planning trajectories that consider future variations. In the proposed method, the probability of maintaining a safe distance from other vehicles is imposed as a chance constraint. In merging scenarios where the behavior of other vehicles is highly variable, the success ratio of merging improved by 30%. Additionally, the performance of the proposed method was evaluated by adapting it to a real vehicle.