Constructive learning approach for minimal intelligent sensing on embedded devices

Abstract

The importance of intelligent control actions has increased and must be addressed within engineering training itineraries. This paper presents part of a practical training itinerary for learning to develop intelligent IoT systems. For this development, TinyML on the Arduino Nano 33 BLE Sense is used as an example of an intelligent embedded system with limited resources. Through three practical modules, the use of an embedded device to perform an intelligent control task is addressed. To achieve this, students progress through three pedagogically validated stages: environmental awareness, one-dimensional signal analysis, and two-dimensional signal analysis. The results demonstrate the feasibility of deploying real-time inference models directly onto microcontrollers, enabling efficient and local automation. This approach provides new teachers with methodological tools to introduce embedded artificial intelligence within the fields of systems engineering and automation.