Mobile Computation in Connected Vehicles

In this chapter, we delve into the transformative role of connected vehicles as dynamic computation platforms, transcending their conventional transportation functions. With the advent of diverse computing hardware such as Graphic Processor Units (GPUs), Digital Signal Processors (DSPs), Field Programmable Gate Arrays (FPGAs), and Application-Specific Integrated Circuits (ASICs), vehicles are now equipped to execute intricate computational tasks. This narrative begins with a historical overview of automotive software and computing systems, highlighting the evolution toward intelligent vehicle applications. Subsequently, it explores the concept of software-defined vehicles, outlining the application landscape and requirements. An in-depth examination of the reference architecture for vehicle computing systems is provided, including state-of-the-art computing systems, real-time operating systems, and middleware solutions like Robot Operating Systems and Autoware. The chapter also addresses the multifaceted challenges associated with vehicle computations, spanning artificial intelligence, vehicle operating systems, hardware costs, and experimental platforms. Through this comprehensive discourse, we illuminate the complex interplay between technology and mobility, charting the path for future innovations in vehicle computing.