Exploring RUST-based AUTOSAR-compliant Operating Systems for Embedded Processors

Electronics and software are essential components in the modern automotive industry, where safety, correctness, and real-time performance are critical requirements. Traditional automotive software relies on C/C++ due to its high performance, but its unrestricted syntax and manual memory management make it highly error-prone, increasing development costs and risks, especially for large projects. Rust is a promising alternative, a mature programming language that offers comparable performance and eliminates entire classes of memory and concurrency errors through strict compile-time rules. The language is gaining traction in the context of embedded systems, supported by an expanding ecosystem of projects and libraries. This thesis aims to demonstrate the feasibility of using Rust in real-time, safety-critical embedded software through the development of a Rust-based operating system compatible with the widely adopted AUTOSAR standard. The project targets a development board featuring the real-time profile Cortex-R52 processor and is also compatible with QEMU. The core component of the system is a priority-based scheduler driven by a hardware timer on the CPU. AUTOSAR objects that define more advanced features such as synchronization mechanisms between tasks and cores, and configurable interrupt routines were implemented, along with a comprehensive set of APIs to expose them to the user code, which remains in C/C++ for compatibility. The final outcome includes build tools, examples, and documentation. It establishes a robust foundation that demonstrates the practical applicability of Rust to real-time, safety-critical embedded systems. By providing a working prototype aligned with AUTOSAR principles, the project offers a starting point that can be extended and refined for future research and industrial adoption.