Adaptive systems research group

The PRELAT project

Aim:
To develop methods for accurate lateral truck positioning

Brief description
In this project, we are using stochastic optimization algorithms and image processing in order to obtain accurate estimates of the lateral position of a truck.

Background and motivation
The longitudinal position (i.e. the position along the road) of a vehicle can typically be estimated quite accurately using GPS (perhaps augmented with odometry to handle GPS signal loss). However, obtaining an accurate estimate of a vehicle's lateral position, i.e. its position in the lane(s), perpendicular to the road tangent, is more challenging, as the (standard) GPS signal is not sufficiently accurate for such measurements. Knowledge of the lateral position is essential for the development of autonomous vehicles.

Detailed description
In this project, we are using landmark-based orientation, based on camera images, in order to estimate the lateral position of a vehicle (and, perhaps, its current heading (yaw) angle as well).

The project started in August 2015. More information will be posted in the Autumn of 2015.

Participants

Mattias Wahde
+ 1 PhD student (hiring in progress)


Last Update: 20150804, 08.00