Extensible Highway Simulation Environment Supporting the Research of Situation Analysis and Behaviour Planning for Autonomous Vehicles

The research on autonomous vehicle technologies is a popular field of interest where evaluating and testing new models and algorithms depends mainly on Software-in-the-Loop simulation solutions. This paper presents a simulator with a simple, two-dimensional viewer whose aim is to help testing the developed functions by providing an efficient way of creating a large amount of traffic scenarios and aiding the labelling process of any collected dataset for training machine learning based algorithms. Our focus was to create a platform, whose functionality can be easily extended with future demands yet keeping its architecture relatively simple. To better understand the context, the concept of the developed system is briefly described as well.