Testing for Vehicle Computing

This chapter serves as a pivotal testing reference for vehicle computing, providing thorough evaluation and testing approaches that are at the heart of connected and autonomous vehicle technologies. The chapter begins by detailing CAVBench, an innovative benchmark suite specifically conceived for vehicle computing systems, offering a closer look at its architecture and the imperative role of benchmarking in enhancing system performance. Subsequently, it examines crucial metrics for computing systems such as accuracy, timeliness, power, cost, reliability, privacy, and security, underlining their relevance in the assessment process. The narrative then transitions into the domain of simulation, contrasting various simulators and emphasizing their significance through case studies like BlueICE and D-STAR. Furthermore, the chapter explores testbeds for vehicle computing, spanning outdoor and indoor environments and delving into the nuances of single-vehicle and multi-vehicle testing, illustrated by the ICAT case study. Finally, the chapter engages in a thoughtful discussion on the challenges posed by experimental platforms and physical worlds coupling, providing a comprehensive overview that is instrumental for academics, industry professionals, and enthusiasts invested in advancing vehicle computing systems.