TinyML applications on embedded systems for industry
In the recent years, artificial intelligence, machine learning and IoT technologies have enabled a great number of industrial applications with profitable results. Predicting the remaining useful life (RUL) of turbofan engines constitutes a successful example of industrial AI, and it has received th...
Κύριος συγγραφέας: | |
---|---|
Άλλοι συγγραφείς: | |
Γλώσσα: | English |
Έκδοση: |
2023
|
Θέματα: | |
Διαθέσιμο Online: | https://hdl.handle.net/10889/24457 |
id |
nemertes-10889-24457 |
---|---|
record_format |
dspace |
spelling |
nemertes-10889-244572023-02-17T04:36:57Z TinyML applications on embedded systems for industry Εφαρμογές TinyML σε ενσωματωμένα συστήματα για τη βιομηχανία Αθανασάκης, Γεώργιος Athanasakis, Georgios Machine learning TinyML Embedded systems Internet of Things (IoT) Industry 4.0 Μηχανική μάθηση Ενσωματωμένα συστήματα Διαδίκτυο των Αντικειμένων In the recent years, artificial intelligence, machine learning and IoT technologies have enabled a great number of industrial applications with profitable results. Predicting the remaining useful life (RUL) of turbofan engines constitutes a successful example of industrial AI, and it has received thorough attention from the researchers worldwide, with numerous novel and effective methods being proposed in the literature. Meanwhile, TinyML is a recent trend that has emerged in the AI field and demonstrates, amongst others, promising potential to break through the existing barriers of trusting and deploying real-time critical industrial AI solutions. In this context, this thesis aims to further contribute to the literature and demonstrate the realization of RUL predictions in the extreme edge via TinyML, using the popular C-MAPSS dataset from National Aeronautics and Space Administration (NASA) Ames Research Center, X-CUBE-AI tool and an STM32 microcontroller for the deployment of ML models. We benchmark different ML algorithms, with a special focus on deep learning algorithms (LSTMs and CNNs). The results indicate that there is potential for deploying machine learning models for RUL prediction in resource-scarce IoT devices, with acceptable accuracy loss, while taking advantage of the benefits TinyML has to offer over cloud-based AI inference. Finally, industrial work areas and production lines can benefit from the implementation of a predictive maintenance solution to increase productivity and decrease downtime. Such an application requires the generation and collection of sensor data, monitoring and processing of this data, as well as running an inference of a machine learning model to make predictions and deliver the necessary output. A PLC is a common device used in industry and can generate similar sensor data. The communication between PLC and STM32 microcontroller is handled by a Raspberry pi which acts as an intermediate. OPC UA protocol is implemented on Raspberry Pi for PLC communication and UART protocol for STM32 communication. 2023-02-16T12:06:44Z 2023-02-16T12:06:44Z 2023-02-16 https://hdl.handle.net/10889/24457 en application/pdf |
institution |
UPatras |
collection |
Nemertes |
language |
English |
topic |
Machine learning TinyML Embedded systems Internet of Things (IoT) Industry 4.0 Μηχανική μάθηση Ενσωματωμένα συστήματα Διαδίκτυο των Αντικειμένων |
spellingShingle |
Machine learning TinyML Embedded systems Internet of Things (IoT) Industry 4.0 Μηχανική μάθηση Ενσωματωμένα συστήματα Διαδίκτυο των Αντικειμένων Αθανασάκης, Γεώργιος TinyML applications on embedded systems for industry |
description |
In the recent years, artificial intelligence, machine learning and IoT technologies have enabled a great number of industrial applications with profitable results. Predicting the remaining useful life (RUL) of turbofan engines constitutes a successful example of industrial AI, and it has received thorough attention from the researchers worldwide, with numerous novel and effective methods being proposed in the literature. Meanwhile, TinyML is a recent trend that has emerged in the AI field and demonstrates, amongst others, promising potential to break through the existing barriers of trusting and deploying real-time critical industrial AI solutions.
In this context, this thesis aims to further contribute to the literature and demonstrate the realization of RUL predictions in the extreme edge via TinyML, using the popular C-MAPSS dataset from National Aeronautics and Space Administration (NASA) Ames Research Center, X-CUBE-AI tool and an STM32 microcontroller for the deployment of ML models. We benchmark different ML algorithms, with a special focus on deep learning algorithms (LSTMs and CNNs). The results indicate that there is potential for deploying machine learning models for RUL prediction in resource-scarce IoT devices, with acceptable accuracy loss, while taking advantage of the benefits TinyML has to offer over cloud-based AI inference.
Finally, industrial work areas and production lines can benefit from the implementation of a predictive maintenance solution to increase productivity and decrease downtime. Such an application requires the generation and collection of sensor data, monitoring and processing of this data, as well as running an inference of a machine learning model to make predictions and deliver the necessary output. A PLC is a common device used in industry and can generate similar sensor data. The communication between PLC and STM32 microcontroller is handled by a Raspberry pi which acts as an intermediate. OPC UA protocol is implemented on Raspberry Pi for PLC communication and UART protocol for STM32 communication. |
author2 |
Athanasakis, Georgios |
author_facet |
Athanasakis, Georgios Αθανασάκης, Γεώργιος |
author |
Αθανασάκης, Γεώργιος |
author_sort |
Αθανασάκης, Γεώργιος |
title |
TinyML applications on embedded systems for industry |
title_short |
TinyML applications on embedded systems for industry |
title_full |
TinyML applications on embedded systems for industry |
title_fullStr |
TinyML applications on embedded systems for industry |
title_full_unstemmed |
TinyML applications on embedded systems for industry |
title_sort |
tinyml applications on embedded systems for industry |
publishDate |
2023 |
url |
https://hdl.handle.net/10889/24457 |
work_keys_str_mv |
AT athanasakēsgeōrgios tinymlapplicationsonembeddedsystemsforindustry AT athanasakēsgeōrgios epharmogestinymlseensōmatōmenasystēmatagiatēbiomēchania |
_version_ |
1771297268709720064 |