Assessment of IoT-driven predictive maintenance strategies for Computed Tomography equipment: a machine learning approach
Predictive maintenance (PdM) identifies the equipment conditions and forecasts when maintenance is required to minimize downtime, whis is crucial for medical equipment. This study developed a machine learning-based PdM for a Computed Tomography (CT) scan machine using Internet of Things (IoT) sensor...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Institute of Electrical and Electronics Engineers
2024
|
| Online Access: | http://psasir.upm.edu.my/id/eprint/114763/ http://psasir.upm.edu.my/id/eprint/114763/1/114763.pdf |
| _version_ | 1848866588313255936 |
|---|---|
| author | Mohd Azrul Shazril, Mohammad Habib Shah Ershad Mashohor, Syamsiah Amran, Mohd Effendi Hafiz, Nur Fatinah Ali, Azizi Mohd Naseri, Mohd Saiful Rasid, Mohd Fadlee A. |
| author_facet | Mohd Azrul Shazril, Mohammad Habib Shah Ershad Mashohor, Syamsiah Amran, Mohd Effendi Hafiz, Nur Fatinah Ali, Azizi Mohd Naseri, Mohd Saiful Rasid, Mohd Fadlee A. |
| author_sort | Mohd Azrul Shazril, Mohammad Habib Shah Ershad |
| building | UPM Institutional Repository |
| collection | Online Access |
| description | Predictive maintenance (PdM) identifies the equipment conditions and forecasts when maintenance is required to minimize downtime, whis is crucial for medical equipment. This study developed a machine learning-based PdM for a Computed Tomography (CT) scan machine using Internet of Things (IoT) sensors to monitor temperature, humidity, current, radiation, and XY-axis acceleration. Data were collected from January to December 2023 at a hospital in the Klang Valley, Malaysia. The readings were preprocessed to follow a normal distribution, representing the typical working conditions of the machine. Owing to limited faulty condition data, synthetic data were generated by expanding the tails of the data distribution and using a Gaussian noise generator. These synthetic data are vital for training robust machine learning models. An artificial neural network (ANN) was designed to predict the machine's breakdown risk using all sensor parameters as inputs. The ANN model achieved an impressive prediction accuracy of 97.58%, proving its relibility in forecasting breakdowns. The model consistently predicted a high breakdown risk in November 2023, which was confirmed by a repair report that indicating maintenance was required in early December 2023. This study demonstrated that integrating IoT sensors with ANN models can significantly enhance the PdM of medical equipment, reduce downtime, and improve operational efficiency. These promising results suggest the potential application of this approach in other critical medical devices. |
| first_indexed | 2025-11-15T14:22:59Z |
| format | Article |
| id | upm-114763 |
| institution | Universiti Putra Malaysia |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-15T14:22:59Z |
| publishDate | 2024 |
| publisher | Institute of Electrical and Electronics Engineers |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | upm-1147632025-01-31T01:07:47Z http://psasir.upm.edu.my/id/eprint/114763/ Assessment of IoT-driven predictive maintenance strategies for Computed Tomography equipment: a machine learning approach Mohd Azrul Shazril, Mohammad Habib Shah Ershad Mashohor, Syamsiah Amran, Mohd Effendi Hafiz, Nur Fatinah Ali, Azizi Mohd Naseri, Mohd Saiful Rasid, Mohd Fadlee A. Predictive maintenance (PdM) identifies the equipment conditions and forecasts when maintenance is required to minimize downtime, whis is crucial for medical equipment. This study developed a machine learning-based PdM for a Computed Tomography (CT) scan machine using Internet of Things (IoT) sensors to monitor temperature, humidity, current, radiation, and XY-axis acceleration. Data were collected from January to December 2023 at a hospital in the Klang Valley, Malaysia. The readings were preprocessed to follow a normal distribution, representing the typical working conditions of the machine. Owing to limited faulty condition data, synthetic data were generated by expanding the tails of the data distribution and using a Gaussian noise generator. These synthetic data are vital for training robust machine learning models. An artificial neural network (ANN) was designed to predict the machine's breakdown risk using all sensor parameters as inputs. The ANN model achieved an impressive prediction accuracy of 97.58%, proving its relibility in forecasting breakdowns. The model consistently predicted a high breakdown risk in November 2023, which was confirmed by a repair report that indicating maintenance was required in early December 2023. This study demonstrated that integrating IoT sensors with ANN models can significantly enhance the PdM of medical equipment, reduce downtime, and improve operational efficiency. These promising results suggest the potential application of this approach in other critical medical devices. Institute of Electrical and Electronics Engineers 2024 Article PeerReviewed text en cc_by_4 http://psasir.upm.edu.my/id/eprint/114763/1/114763.pdf Mohd Azrul Shazril, Mohammad Habib Shah Ershad and Mashohor, Syamsiah and Amran, Mohd Effendi and Hafiz, Nur Fatinah and Ali, Azizi Mohd and Naseri, Mohd Saiful and Rasid, Mohd Fadlee A. (2024) Assessment of IoT-driven predictive maintenance strategies for Computed Tomography equipment: a machine learning approach. IEEE Access, 12. pp. 195505-195515. ISSN 2169-3536; eISSN: 2169-3536 https://ieeexplore.ieee.org/document/10804159/ 10.1109/ACCESS.2024.3518516 |
| spellingShingle | Mohd Azrul Shazril, Mohammad Habib Shah Ershad Mashohor, Syamsiah Amran, Mohd Effendi Hafiz, Nur Fatinah Ali, Azizi Mohd Naseri, Mohd Saiful Rasid, Mohd Fadlee A. Assessment of IoT-driven predictive maintenance strategies for Computed Tomography equipment: a machine learning approach |
| title | Assessment of IoT-driven predictive maintenance strategies for Computed Tomography equipment: a machine learning approach |
| title_full | Assessment of IoT-driven predictive maintenance strategies for Computed Tomography equipment: a machine learning approach |
| title_fullStr | Assessment of IoT-driven predictive maintenance strategies for Computed Tomography equipment: a machine learning approach |
| title_full_unstemmed | Assessment of IoT-driven predictive maintenance strategies for Computed Tomography equipment: a machine learning approach |
| title_short | Assessment of IoT-driven predictive maintenance strategies for Computed Tomography equipment: a machine learning approach |
| title_sort | assessment of iot-driven predictive maintenance strategies for computed tomography equipment: a machine learning approach |
| url | http://psasir.upm.edu.my/id/eprint/114763/ http://psasir.upm.edu.my/id/eprint/114763/ http://psasir.upm.edu.my/id/eprint/114763/ http://psasir.upm.edu.my/id/eprint/114763/1/114763.pdf |