AI coding agents implemented six ways to update a Matter device on real hardware (groundrun.io)

AI coding agents successfully implemented six different methods for Over-The-Air (OTA) updates on a Matter smart home device using the ESP32-H2 chip, completing the task in 29.7 hours. This project is significant for the AI/ML community as it demonstrates the capabilities of AI in automating complex coding processes, especially for critical functions like OTA updates, which are essential for maintaining the security and functionality of connected devices. The experiment not only validated different OTA mechanisms but also discovered a bug in the Matter code, highlighting the importance of continuous testing and refinement in software development. The agents utilized a Groundrun hardware setup, executing a series of planned tasks that included full firmware updates with varying block sizes and a delta-encoded update that efficiently sent only the differences in the firmware. Results indicated that smaller block sizes increased transfer times due to the required stop-and-wait protocol in data transmission. This exploration of OTA mechanisms and performance metrics is crucial as it lays the groundwork for reliable smart home device updates, reflecting the regulatory pressures for security compliance, such as the EU’s Cyber Resilience Act. This work underscores the growing intersection between AI-driven software development and everyday technology.