Evaluating the Impact of RTOS and Edge AI on the Power Consumption of STM32-Based IoT Nodes Abstract As Internet of Things (IoT) nodes become increasingly integrated into complex industrial systems, minimizing their systemic power consumption remains a paramount design challenge. A primary factor dictating this energy footprint is the firmware execution paradigm, particularly when managing high-frequency sensors, advanced Edge Artificial Intelligence (AI), and Long Range (LoRa) telemetry. While existing literature frequently relies on static datasheet values to estimate energy usage, these theoretical calculations fail to capture the dynamic, microsecond-level load transients of real-world operation, leading to highly inaccurate battery life projections.