Four Decades of Automated Driving R&D: Evolution, Challenges, Future Directions

Abstract

Enhancing driver comfort and road safety has consistently driven the development of intelligent road vehicles. These vehicles leverage information and communication technology (ICT) to support advanced driver assistance and automated driving functions within the domain of intelligent transportation systems (ITS). Over four decades, these technologies have transitioned from basic navigation and assistance features toward highly complex automated driving systems (ADS). The growing complexity of ADS, particularly the integration of multi-sensor perception and data-driven algorithms, challenges traditional approaches to vehicle verification and validation (V&V). Unlike conventional automotive systems, based largely on deterministic physical behavior, modern intelligent vehicles exhibit software-intensive and partially non-deterministic characteristics that require new safety assurance methods. This paper presents a technical review of the evolution of technologies, architectures and validation approaches supporting intelligent vehicles. By adopting a holistic perspective on safety assessment and V&V throughout the development and deployment lifecycle of ADS, with primary focus on the dynamic driving task (DDT), the paper synthesizes the lessons learned from their evolution, identifies critical development gaps, analyzes current challenges, and outlines future research directions to support their safe deployment.