Open AccessAn industrial IoT edge node for buffer level detection in a cardboard production line
Author(s) -
George Avramidis,
Dimitris Karampatzakis
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.179
H-Index - 26
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1032/1/012014
Subject(s) - computer science , implementation , node (physics) , edge computing , overhead (engineering) , enhanced data rates for gsm evolution , internet of things , key (lock) , the internet , bridge (graph theory) , embedded system , artificial intelligence , operating system , engineering , software engineering , medicine , structural engineering
Computer Science and Internet have evolved rapidly the last decades and equally impressive is the evolution of the Industrial Internet of Things technology into factories’ shop floors. Among other technologies: modern CPU Architectures, Edge Computing, Deep Learning, Computer Vision, and Low Power Wide Area Networks, are playing a key role in this new competitive environment. In this paper, we present an Industrial IoT Edge Node for level detection on an overhead bridge conveyor (buffer) which is part of a 5-ply corrugated cardboard production line. We focused on the Edge Node and the development of the system was accomplished by using state of the art technologies from disciplines of computer vision and deep learning. We present two implementations using contour detection and CNN techniques. Finally, we implemented a LoRaWAN solution in the IIoT node to send alert messages to the control room. Experimental results are presented for the proposed system implementations.