The MS module ( Microprocessors Systems ) provides an introduction to the development of Embedded Linux-based applications for mobile platforms with ARM processors (ARM A53/72,.., ARM Cortex M0/M33). This module is a gateway between ECU (Electronic Control Unit) systems equipped with sensors/actuators and the universe of applications based on CLOUD and Machine Learning resources.
The central, innovative element consists in expanding the Unix/Linux universe with that of sensors and actuators connected in the real world which, together with the Internet and GSM networks, allow the creation of a true digital nervous system with a planetary dimension (Internet of Things).
Acquiring knowledge and developing skills to apply it in concrete applications involves "learning by doing". In this sense, it is necessary to use a TinyML Learning Kit based on a set of hardware components and associated software resources, capable of allowing the replication of the proposed set of educational experiments (Fig.1) and to allow the running of inferences made with TensorFlow.
Fig. 1 TinyML Kit Components
The core consists of a development platform with ARM processors capable of running a Linux variant (Debian, Arch, Ubuntu, openSUSE, CentOS etc). Development systems from the Raspberry Pi zero/5 family are used, but relatively equivalent platforms can also be used such as NVIDIA Jetson Nano, NXP i.MX8M, Intel Galileo-Edison, etc.
The Python language is the main tool used for application development. The kit also contains a breadboard, LEDs, optotriac, I2C HD4478 display , servo, RGB LED, ultrasonic distance sensor type HC-SR04, ADC converter type ADS1115, Nokia LCD, USB-RS223 converter, GSM module, Pi video camera , temperature sensors, acceleration sensors, relays, measuring device, letcon and Raspberry Pi Pico 2 for real-time/ML applications.
The lab experiments are designed as 8-dimensional training vector models, accompanied by a complete set of information so that it can be replicated on the equipment in your own garage.
The courses and lab experiments aim to develop robust solutions, with students subsequently identifying real-world problems for which they are valid (totally/partially).
In this way, the classic system is avoided in which, as a rule, problems are given and solutions are sought, which are, most of the time, difficult to achieve in limited time constraints.
The online support, created according to the principles of MIT-OCW, ensures the exploration of materials anytime, from anywhere, by anyone interested.
The SM module (argued with chatGPT/GEMINI) is especially aimed at students passionate about technology and concerned with the development of new products/services using the most fascinating universe of computer engineering - Embedded Machine Learning - TinyML .
Motto SM:
"You don't learn to walk by following rules. You learn by doing, and by falling over."
Richard Branson