Vehicle-in-Virtual-Environment (VVE) Based Autonomous Driving Function Development and Evaluation Methodology for Vulnerable Road User Safety

TL;DR

This paper introduces a Vehicle-in-Virtual-Environment (VVE) approach that enhances the development and testing of autonomous driving functions, especially for vulnerable road user safety, by combining real vehicles with virtual scenarios for safer, more efficient evaluation.

Contribution

It proposes a new testing pipeline integrating MIL, HIL, and VVE methods for comprehensive autonomous vehicle development and evaluation.

Findings

VVE enables safe, realistic testing of traffic scenarios involving vulnerable road users.

The pipeline improves safety and efficiency in autonomous driving system development.

Demonstrated with a path-tracking algorithm incorporating deep reinforcement learning for collision avoidance.

Abstract

Traditional methods for developing and evaluating autonomous driving functions, such as model-in-the-loop (MIL) and hardware-in-the-loop (HIL) simulations, heavily depend on the accuracy of simulated vehicle models and human factors, especially for vulnerable road user safety systems. Continuation of development during public road deployment forces other road users including vulnerable ones to involuntarily participate in the development process, leading to safety risks, inefficiencies, and a decline in public trust. To address these deficiencies, the Vehicle-in-Virtual-Environment (VVE) method was proposed as a safer, more efficient, and cost-effective solution for developing and testing connected and autonomous driving technologies by operating the real vehicle and multiple other actors like vulnerable road users in different test areas while being immersed within the same highly realistic virtual environment. This VVE approach synchronizes real-world vehicle and vulnerable road user motion within the same virtual scenario, enabling the safe and realistic testing of various traffic situations in a safe and repeatable manner. In this paper, we propose a new testing pipeline that sequentially integrates MIL, HIL, and VVE methods to comprehensively develop and evaluate autonomous driving functions. The effectiveness of this testing pipeline will be demonstrated using an autonomous driving path-tracking algorithm with local deep reinforcement learning modification for vulnerable road user collision avoidance.