Architectures and strategies for low power optimization in the ESP32 ecosystem: A systematic review of the literature

Abstract

The objective is to identify the optimal configurations and strategies for low power optimization in the ESP32 series microcontroller ecosystem, addressing the need to balance computing power with energy autonomy in autonomous systems. The methodology employed consisted of a structured analysis based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) protocol, selecting twenty-eight technical studies. Key findings reveal that efficiency depends not only on deep sleep states but also on a comprehensive, hardware-sensitive co-design paradigm. The reviewed studies demonstrate that the use of native compiled languages achieves execution speeds up to fifteen times faster than interpreted options, while the application of eight-bit quantization techniques reports reductions of up to seventy-three percent in memory footprint. Furthermore, the hybridization of short- and long-range communication protocols optimizes operational resilience in remote environments. It is concluded that these strategies transform the device from a prototyping tool into a robust scientific instrument for edge artificial intelligence. Practical implications include the development of medical biosensors with months of autonomy and low-cost industrial monitoring networks, democratizing access to advanced digital technologies in critical sectors and regions with limited infrastructure.