TinyML-Based Edge Intelligence Architecture for Industrial IoT Systems Using LoRa Networks

• Grant number: 26/18105-3
• Support Opportunities: Research Grants - Regular Innovation Grant - AIR
• Start date: October 01, 2026
• End date: March 31, 2028
• Field of knowledge: Engineering - Electrical Engineering - Industrial Electronics, Electronic Systems and Controls
• Principal Investigator: Renan Rodrigues Mendes
• Grantee: Renan Rodrigues Mendes
• Host Institution: FAC TECNOLOGIA DIADEMA LUIGI PAPAIZ
• Associated researchers: Claudio Rodrigo Torres
• Associated scholarship(s): 26/24053-6 - TinyML-Based Edge Intelligence Architecture for Industrial IoT Systems Using LoRa Networks, BP.IT

Abstract

The growing adoption of the Industrial Internet of Things (IIoT) has expanded the use of intelligent sensors for monitoring industrial processes. However, many existing architectures still rely on cloud-centric processing, resulting in increased energy consumption, higher latency, and greater network traffic, particularly in low-power communication technologies such as LoRa. This project proposes a TinyML-based Edge Intelligence architecture integrated with LoRa communication, in which machine learning inference is performed directly on embedded devices, enabling only relevant events to be transmitted to the server. This approach reduces data transmission, optimizes energy consumption, and improves system scalability. As a case study, a prototype will be developed for monitoring industrial forklifts using ESP32-based embedded devices, sensors, and TinyML models to detect critical operational events. The proposed architecture will be modular and reusable, allowing its application to other IIoT scenarios, including smart agriculture, autonomous vehicles, and environmental monitoring. The expected outcome is to advance Edge AI solutions for industrial systems, supporting Industry 4.0 and the Digital Transformation through energy-efficient, distributed intelligent processing. (AU)