5G-Based Multi-Sensor Platform for Monitoring of Workpieces and Machines: Prototype Hardware Design and Firmware

Abstract

In this paper, we introduce a 5G-based multi-sensor platform for monitoring workpieces and machines. The prototype is realized within the EU-funded 5G-SMART project, which aims to enable smart manufacturing through 5G, demonstrating and validating new generation network technology in industrial processes. There are already state-of-the-art solutions, but with drawbacks such as limited flexibility, brief real-time capability, and sensors aimed at single applications. The 5G-SMART multi-sensor platform is designed to overcome these points and meet the requirements of Industry 4.0. The device is equipped with different sensors to acquire multiple data from workpieces and machines of the shop floor, wirelessly connected by 5G to the factory cloud. A hardware design description of the prototype is provided, focusing on the electronic components and their interaction with the microcontroller. Verification of the correct functioning of the board is given, with a basic library for the main peripherals used as a basis for the final firmware.

Architecture for Edge-Based Predictive Maintenance of Machines Using Federated Learning and Multi Sensor Platforms

“…The architecture system described was developed using an open-source framework, namely IBM Federated Learning (IBMFL) [20], as core module for the FL platform. From the hardware side, the MSP [21] described is developed using the STM32H7 series microcontroller. The programming language is Embedded C using the STM32 cube IDE.…”