Sistem Prediksi Suhu, Kelembapan, dan Tekanan Udara Pada Ruang Pengemasan Vaksin Menggunakan Metode Long Short Term Memory

Fattah, Muhammad Awalu (2024) Sistem Prediksi Suhu, Kelembapan, dan Tekanan Udara Pada Ruang Pengemasan Vaksin Menggunakan Metode Long Short Term Memory. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2040201072-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (7MB) | Request a copy

Abstract

PT Bio Farma berupaya memastikan bahwa kondisi lingkungan di ruang pengemasan vaksin memenuhi standar dan regulasi Cara Pembuatan Obat yang Baik (CPOB), karena faktor seperti suhu, kelembapan, dan tekanan udara berpengaruh langsung terhadap stabilitas serta keefektifan vaksin. Saat ini, Ruang Pengemasan Vaksin telah dilengkapi dengan sistem pendingin Air Handling Unit (AHU) yang terintegrasi dengan Building Management System (BMS). Namun, sistem tersebut belum mampu memprediksi anomali atau menganalisis kondisi di masa lalu untuk memproyeksikan situasi di masa mendatang sebagai bahan evaluasi dalam produksi. Proyek akhir ini bertujuan untuk mengatasi kekurangan tersebut dengan mengimplementasikan sistem prediksi suhu, kelembapan, dan tekanan udara di ruang pengemasan vaksin. Sistem ini tidak hanya memonitor kondisi lingkungan secara real-time tetapi juga memungkinkan identifikasi dini terhadap potensi masalah. Dengan demikian, risiko ketidaksesuaian dapat diminimalkan dan efisiensi produksi vaksin dapat ditingkatkan. Menggunakan algoritma Long Short Term Memory (LSTM), sistem ini memprediksi kondisi suhu, kelembapan, dan tekanan udara di masa depan. Integrasi dengan sistem IoT menjadi solusi untuk pemberitahuan dini terkait kemungkinan downtime dan kegagalan tak terduga dalam proses produksi vaksin. Sistem ini dirancang menggunakan Raspberry Pi 4B+ dan sensor BME280, yang diintegrasikan dengan InfluxDB dan Grafana. Hasil uji menunjukkan performa prediksi yang baik, dengan konfigurasi hyperparameter LSTM yang terdiri dari 50 neuron dan 100 epoch menggunakan optimizer Adam. Sistem ini menghasilkan nilai Mean Squared Error (MSE) sebesar 0,000819, Mean Absolute Error (MAE) sebesar 0,02045, dan Mean Absolute Percentage Error (MAPE) sebesar 0,0398.
=======================================================================================================================================
PT Bio Farma strives to ensure that the environmental conditions in the vaccine packaging room comply with the Good Manufacturing Practice (GMP) standards and regulations, as factors such as temperature, humidity, and air pressure directly affect the stability and efficacy of vaccines. Currently, the Vaccine Packaging Room is equipped with an Air Handling Unit (AHU) cooling system integrated with a Building Management System (BMS). However, this system is not yet capable of predicting anomalies or analyzing past conditions to project future scenarios for production evaluation. This final project aims to address these shortcomings by implementing a predictive system for temperature, humidity, and air pressure in the vaccine packaging room. This system not only monitors environmental conditions in real-time but also enables early identification of potential issues. Consequently, the risk of non-compliance can be minimized, and the overall efficiency of vaccine production can be enhanced. Using the Long Short Term Memory (LSTM) algorithm, this system predicts future temperature, humidity, and air pressure conditions. Integration with the Internet of Things (IoT) provides an alternative solution for early warnings related to potential downtime and unexpected failures in the vaccine production process. The system is designed using a Raspberry Pi 4B+ and BME280 sensor, integrated with InfluxDB and Grafana. Test results demonstrate good predictive performance, with an LSTM hyperparameter configuration consisting of 50 neurons and 100 epochs using the Adam optimizer. The system achieved a Mean Squared Error (MSE) of 0.000819, a Mean Absolute Error (MAE) of 0.02045, and a Mean Absolute Percentage Error (MAPE) of 0.0398.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	IoT, LSTM, Ruang Pengemasan, Sistem Prediksi, Vaksin, IoT, LSTM, Packaging Room, Predictive System, Vaccine
Subjects:	T Technology > T Technology (General) > T174 Technological forecasting T Technology > T Technology (General) > T385 Visualization--Technique T Technology > T Technology (General) > T57.5 Data Processing T Technology > TJ Mechanical engineering and machinery > TJ213 Automatic control. T Technology > TJ Mechanical engineering and machinery > TJ217 Adaptive control systems T Technology > TJ Mechanical engineering and machinery > TJ217.6 Predictive Control
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Muhammad Awalu Fattah
Date Deposited:	26 Aug 2024 00:40
Last Modified:	26 Aug 2024 00:40
URI:	http://repository.its.ac.id/id/eprint/115525

Actions (login required)

View Item