Merit-Based Motion Planning for Autonomous Vehicles in Urban Scenarios.

Juan Medina-Lee,Jorge Villagra,Jorge Godoy,Antonio Artuñedo

doi:10.3390/s21113755

Abstract

Safe and adaptable motion planning for autonomous vehicles remains an open problem in urban environments, where the variability of situations and behaviors may become intractable using rule-based approaches. This work proposes a use-case-independent motion planning algorithm that generates a set of possible trajectories and selects the best of them according to a merit function that combines longitudinal comfort, lateral comfort, safety and utility criteria. The system was tested in urban scenarios on simulated and real environments, and the results show that different driving styles can be achieved according to the priorities set in the merit function, always meeting safety and comfort parameters imposed by design.

Highlights

Automated Driving Functions (ADF) are progressing at a vertiginous pace
End-to-end solutions (e.g., [4]), enabled by deep and imitation learning, are achieving impressive performance. All these strategies may be successful in many cases, one drawback is that they are designed for specific Operational Design Domains (ODD) and sometimes produce inexplicable behaviors, which makes it hard to scale them to the complexity of real-world urban driving
The average jerk values were higher in the real environment, which was the reason for this the noisy signals of the accelerometers on board the automated vehicle

Summary

Introduction

Operational Design Domains (ODD), but there are still open problems for a safe navigation in urban environments. In these contexts, decision-making is significantly challenging, as the artificial system must properly interact with a diversity of traffic participants and consider sensors limitations under very different driving situations. End-to-end solutions (e.g., [4]), enabled by deep and imitation learning, are achieving impressive performance All these strategies may be successful in many cases, one drawback is that they are designed for specific ODDs and sometimes produce inexplicable behaviors, which makes it hard to scale them to the complexity of real-world urban driving

Objectives

Methods

Results

Conclusion