Optimizing Lstm-Based Temperature Prediction Algorithm for Embedded System Deployment

This paper presents a flexible Industrial Internet of Things (IIoT) infrastructure model, highlighting the integration of the Long Short-Term Memory (LSTM) algorithm for predictive analysis. A key component of this model is the Concentrator, a fog-computing local hub that provides a sandbox environm...

Full description

Saved in:
Bibliographic Details
Published in2024 IEEE 29th International Conference on Emerging Technologies and Factory Automation (ETFA) pp. 01 - 07
Main Authors d'Agostino, Pietro, Violante, Massimo, Macario, Gianpaolo
Format Conference Proceeding
LanguageEnglish
Published IEEE 10.09.2024
Subjects
Deep learning
Embedded systems
Fog computing
Hardware
Industrial Internet of Things
Industry 4.0
Long short term memory
LSTM
Machine learning algorithms
Prediction algorithms
Predictive maintenance
Predictive models
Raspberry Pi
Temperature prediction
Training
Transfer learning
Virtualization
Online AccessGet full text

Cover

More Information
Summary:This paper presents a flexible Industrial Internet of Things (IIoT) infrastructure model, highlighting the integration of the Long Short-Term Memory (LSTM) algorithm for predictive analysis. A key component of this model is the Concentrator, a fog-computing local hub that provides a sandbox environment for third-party developments. Within this framework, clients can collect data using mechanisms provided by Original Equipment Manufacturers (OEMs), such as AROL Closure Systems, while independently developing proprietary algorithms. This approach eliminates the need for direct interaction with OEMs. The paper explores the use of the LSTM algorithm to develop a predictor for analyzing machine temperature behavior, allowing for the anticipation of potential faults.
ISSN:1946-0759
DOI:10.1109/ETFA61755.2024.10711142