Perancangan Model Sistem Monitoring Mesin Kapal Berbasis Internet Of Things (IoT): Arsitektur 4-Tier Dengan Deteksi Anomali Berbasis Machine Learning Dan Analisis Keandalan Sistem
DOI:
https://doi.org/10.55642/eatij.v6i03.1277Keywords:
IoT kapal; monitoring mesin; ESP32; MQTT; deteksi anomali; LoRaWAN; predictive maintenance; edge computingAbstract
Kegagalan mesin kapal yang tidak terdeteksi dini merupakan penyebab 28% insiden di laut menurut data IACS (2022), dengan biaya perbaikan darurat rata-rata 3–8× lebih tinggi dibandingkan perawatan terencana. Penelitian ini menyajikan perancangan dan implementasi prototipe sistem monitoring mesin kapal berbasis Internet of Things (IoT) dengan arsitektur 4-tier yang mengintegrasikan lapisan persepsi sensor, jaringan komunikasi ganda (WiFi 802.11n + LoRaWAN sebagai backup), edge computing, dan cloud analytics. Kebaruan penelitian mencakup: (1) implementasi algoritma deteksi anomali Isolation Forest + Gaussian Mixture Model (GMM) pada edge node (Raspberry Pi 4) yang mampu memproses data 8 sensor secara simultan dengan latensi median 11,5 ms; (2) strategi komunikasi adaptif dual-channel dengan failover otomatis dari WiFi ke LoRaWAN; (3) validasi akurasi sensor terhadap instrumen referensi kalibrasi (RMSE suhu = 0,724°C, RMSE tekanan = 0,048 bar); dan (4) analisis komprehensif keandalan sistem selama 40 minggu pengujian lapangan (rata-rata uptime = 99,12%). Prototipe diuji pada kapal motor diesel dengan 8 parameter mesin: suhu, tekanan oli, getaran, RPM, flow bahan bakar, kualitas oli, dan knock sensor. Sistem berhasil mendeteksi 3 insiden anomali nyata dengan precision = 91,3%, recall = 87,6%, F1 = 89,4%, dan AUC-ROC = 0,953. Analisis biaya-manfaat menunjukkan payback period 1,2 tahun dengan penghematan bersih IDR 52 juta/tahun.
Downloads
References
IACS, 2022, Casualty Statistics and Investigations: Annual Report 2022, International Association of Classification Societies, London.
KNKT, 2022, Laporan Tahunan Investigasi Kecelakaan Transportasi Laut 2022, Komite Nasional Keselamatan Transportasi, Jakarta.
Wollschlaeger, M., Sauter, T., and Jasperneite, J., 2017, The future of industrial communication: Automation networks in the era of the Internet of Things and Industry 4.0, IEEE Industrial Electronics Magazine, vol. 11, no. 1, pp. 17–27.
McKinsey Global Institute, 2021, The Internet of Things: Catching Up to an Accelerating Opportunity, McKinsey & Company, New York.
Lee, I. and Lee, K., 2015, The Internet of Things (IoT): Applications, investments, and challenges for enterprises, Business Horizons, vol. 58, no. 4, pp. 431–440.
Progoulakis, I., Rohmeyer, P., and Nikitakos, N., 2021, Towards digital engineering for marine and offshore systems, Ocean Engineering, vol. 234, pp. 109–120.
Adland, R., Cariou, P., Jia, H., and Wolff, F. C., 2021, The energy efficiency effects of shaft power limitations in bulk shipping, Transportation Research Part D, vol. 93, pp. 102–115.
Industrial Internet Consortium, 2019, Industrial Internet Reference Architecture, Version 1.9, IIC, Boston.
Shi, W., Cao, J., Zhang, Q., Li, Y., and Xu, L., 2016, Edge computing: Vision and challenges, IEEE Internet of Things Journal, vol. 3, no. 5, pp. 637–646.
Banks, A. and Gupta, R., 2014, MQTT Version 3.1.1, OASIS Standard, https://mqtt.org/mqtt-specification/.
Finneran, T., 2018, Marine Diesel Basics 1: Maintenance, Lay-Up, Winter Protection, Tropical Storage, Spring Commissioning, Fernhurst Books, Brighton.
ISO, 2009, ISO 10816-3:2009 — Mechanical Vibration: Evaluation of Machine Vibration by Measurements on Non-Rotating Parts, International Organization for Standardization, Geneva.
Petajajarvi, J., Mikhaylov, K., Roivainen, A., Hanninen, T., and Pettissalo, M., 2015, On the coverage of LPWANs: range evaluation and channel attenuation model for LoRa technology, in Proceedings of the 14th International Conference on ITS Telecommunications, Copenhagen, pp. 55–59.
Semtech, 2019, AN1200.22 – LoRa Modulation Basics, Application Note, Semtech Corporation, Camarillo, CA.
Liu, F. T., Ting, K. M., and Zhou, Z. H., 2008, Isolation forest, in Proceedings of the 8th IEEE International Conference on Data Mining (ICDM 2008), Pisa, pp. 413–422.
