An Empirical Analysis and Benchmarking Framework for FreeRTOS on the Arduino Uno

Abstract

The scalability of FreeRTOS makes it a popular choice for embedded systems, yet its behavior on processors with severe memory constraints remains poorly quantified. This study conducts a rigorous empirical evaluation of FreeRTOS on the Arduino Uno platform, based on the ATmega328P microcontroller with only 2KB of SRAM. We introduce a systematic benchmarking framework designed to probe the kernel's operational limits by measuring task capacity, stack depth requirements, and scheduler overhead under controlled stress conditions. Our findings reveal that heap fragmentation is the primary determinant of task capacity and precisely quantify the performance trade-offs associated with kernel configuration. The results provide a set of validated, practical guidelines for developers, enabling more reliable design choices for deploying RTOS-based solutions on resource-limited hardware commonly found in cost-sensitive IoT applications.