Afrizal, Aldi (2025) Implementasi Generative Artificial Intelligence dalam Pengambilan Keputusan Maintenance menggunakan Metode Large Language Model. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5010211066-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5010211066-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (8MB) | Request a copy |
Abstract
Computerized Maintenance Management System (CMMS) umumnya hanya dimanfaatkan sebagai alat pencatatan aktivitas pemeliharaan tanpa digunakan secara optimal sebagai dasar pengambilan keputusan. Padahal, data historis gangguan dan pekerjaan pemeliharaan yang tercatat dalam CMMS menyimpan potensi besar untuk membentuk basis pengetahuan yang dapat mendukung knowledge management dalam sistem pemeliharaan. Penelitian ini mengusulkan pendekatan Generative AI dengan memanfaatkan language model seperti LLaMA 2–7B, GPT Neo 1.3B, dan Flan-T5 untuk menghasilkan rekomendasi tindakan pemeliharaan berdasarkan data historis. Dataset yang digunakan mencakup laporan Failure Mode, Effects, and Criticality Analysis (FMECA) dari platform CMMS SAP dan JDE dalam rentang waktu 2016 hingga 2023. Ketiga model diintegrasikan ke dalam sistem Retrieval-Augmented Generation (RAG) berbasis gradio chatbot guna menyediakan saran tindakan pemeliharaan yang kontekstual, relevan, dan mendukung pengambilan keputusan secara cepat dan berbasis data di lingkungan industri. Evaluasi performa model dilakukan menggunakan metrik BLEU, ROUGE, dan BERTScore, serta dilengkapi perhitungan confidence interval pada tingkat kepercayaan 95% untuk menilai konsistensi serta signifikansi hasil. Hasil menunjukkan bahwa LLaMA 2–7B memberikan performa terbaik dengan skor BLEU sebesar 0.16, ROUGE-1 F sebesar 0.19, dan BERTScore sebesar 0.52, serta memiliki rentang confidence interval yang tidak bersinggungan dengan model lainnya, mengindikasikan keunggulan yang signifikan secara statistik.
=================================================================================================================================
Computerized Maintenance Management Systems (CMMS) are generally used merely as tools for recording maintenance activities, without being optimally utilized as a basis for decision-making. In fact, historical data on failures and maintenance tasks recorded in CMMS hold significant potential to form a knowledge base that can support knowledge management within maintenance systems. This study proposes a Generative AI approach by utilizing language models such as LLaMA 2–7B, GPT Neo 1.3B, and Flan-T5 to generate maintenance action recommendations based on historical data. The dataset used includes Failure Mode, Effects, and Criticality Analysis (FMECA) reports from CMMS platforms SAP and JDE, covering the period from 2016 to 2023. All three models are integrated into a RetrievalAugmented Generation (RAG) system deployed through a Gradio-based chatbot to provide contextual and relevant maintenance suggestions that support fast, data-driven decision-making in industrial environments. The performance of each model is evaluated using automated metrics such as BLEU, ROUGE, and BERTScore, and further assessed through confidence interval calculations at a 95% confidence level to examine the consistency and statistical significance of the results. The findings show that LLaMA 2–7B achieved the best performance, with a BLEU score of 0.16, ROUGE-1 F of 0.19, and BERTScore of 0.52, along with
confidence intervals that do not overlap with those of the other models, indicating a statistically significant advantage.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | CMMS, Knowledge Management, Generative AI, Decision Making |
| Subjects: | T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T58.62 Decision support systems |
| Divisions: | Faculty of Industrial Technology and Systems Engineering (INDSYS) > Industrial Engineering > 26201-(S1) Undergraduate Thesis |
| Depositing User: | Aldi Afrizal |
| Date Deposited: | 28 Jul 2025 10:03 |
| Last Modified: | 28 Jul 2025 10:03 |
| URI: | http://repository.its.ac.id/id/eprint/122348 |
Available Versions of this Item
- Implementasi Generative Artificial Intelligence dalam Pengambilan Keputusan Maintenance menggunakan Metode Large Language Model. (deposited 28 Jul 2025 10:03) [Currently Displayed]
Actions (login required)
 |
View Item |