Kazuya Tamura

Sensor and data fusion design and evaluation with a virtual environment simulator

A virtual environment simulation is described, which is useful for the evaluation and testing of on-vehicle ITS sensors and control suites. It is capable of 3D environment simulation and scene generation for use as input to image processing systems, the simulation of complex six degree of freedom (X, Y, Z, roll, pitch, yaw) vehicle models, the generation of ground truth and simulated sensor outputs (i.e., inertial sensors, GPS, vehicle speed), and closed loop testing. The simulator can accept road geometry and terrain data as derived from GIS databases. It is implemented as a modular and extensible software simulation environment using distributed, multiprocessing, networked software modules. Use of the simulator is illustrated by the parameter optimization and evaluation of a sensor fusion algorithm involving GPS, inertial sensors, and a map database.

Autonomous vehicle control system of ICVS City Pal: electrical tow-bar function

A platooning system, in which vehicles follow a lead vehicle without the road infrastructure, by combining laser radar sensing systems and inter-vehicle communication and navigation system has been developed. It tracks automatically the race line of a lead vehicle. The maximum of five vehicles including a lead vehicle with the minimum inter-vehicle distance of 2 meters ran at the maximum speed of 100 km/h. The accuracy of tracking control was within /spl plusmn/30 cm for the longitudinal direction and /spl plusmn/15 cm for the lateral direction.

A study of self-reliant cornering speed control system

A major cause of accidents during cornering is the excessive vehicle speed due to the drivers incorrect recognition or misjudgement of the curvature of the road, the vehicle performance and the road conditions. We are now studying a self-reliant cornering vehicle speed control system without using outside facilities along the road. This paper presents the base technology of such a system which predicts the curvature of the oncoming road based on the map information from the car navigation system which has already been installed in production vehicles in Japan.<<ETX>>