What is it about?
The underlying idea is that on intersection vehicles no longer follow the lanes, but curve around each other. So, trajectories of the autonomous vehicle and the surrounding overlap in space, but better not overlap in time. By estimating the speed of the other vehicles no an absolute prediction could be made, because the other vehicles could slow down or speed up, but a risk assesment could be made. This risk assesment could be used to select the safest speed of the autonomous vehicle to cross the intersection.
Featured Image
Photo by Erik Mclean on Unsplash
Why is it important?
When more and more vehicles become autonomous, they could connect wirelessly and start communicating their intentions. This will give no guarantee, but better predictions of the other vehicles behaviors would make intersect crossings go more smoothly.
Perspectives
This paper is based on the CARLA simulation, which I used before, and is very realistic to automate a single vehicle. Yet, with the SUMO extension, the behaviors of other vehicles could also realisticly simulated, which allows to study the dynamics of crossing an intersection.
Dr. Arnoud Visser
Universiteit van Amsterdam
Read the Original
This page is a summary of: Segment-Based Trajectory Prediction and Risk Assessment for RSU-assisted CAVs at Signalized Intersections, IEEE Transactions on Intelligent Vehicles, January 2024, Institute of Electrical & Electronics Engineers (IEEE),
DOI: 10.1109/tiv.2024.3414198.
You can read the full text:
Contributors
The following have contributed to this page
