Claim Missing Document
Check
Articles

Found 2 Documents
Search

Designing an IoT-Based EMIS using Linear Regression and CUSUM for Real-Time Anomaly Detection in Pharmaceutical Industry Freska Lionia Darlion; Muhammad Isa Lufti
Jurnal PASTI (Penelitian dan Aplikasi Sistem dan Teknik Industri) Vol 19, No 3 (2025): Jurnal PASTI
Publisher : Universitas Mercu Buana

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22441/0.22441/10.22441/pasti.2025.v19i3.011

Abstract

The pharmaceutical industry is a high energy-intensity sector requiring strict operational stability. However, many companies still rely on manual energy monitoring methods, leading to information latency. A case study at a multinational pharmaceutical company in Indonesia revealed a baseload inefficiency of 14.3 kW during non-operational hours, which remained undetected due to an ad-hoc energy management system. This study aims to design an IoT-based Energy Management Information System (EMIS) architecture to transform the energy management business process from reactive to proactive-predictive. The study utilizes secondary data from the 2025 Energy Audit Report. The system design integrates a linear regression model (R²=0.80) for Energy Performance Indicators (EnPI) determination and the Cumulative Sum (CUSUM) algorithm for real-time anomaly detection. Investment feasibility is evaluated using techno-economic analysis. The implementation of EMIS requires an investment of IDR 225,000,000 with potential annual energy cost savings of IDR 44,172,687. Although the Simple Payback Period (SPP) is 5.1 years, the project is considered feasible due to its strategic value in data transparency, operational risk mitigation, and ISO 50001 compliance. Furthermore, this digital transformation supports the achievement of Sustainable Development Goals (SDG 7, 9, and 12) by promoting energy efficiency and responsible industrial consumption. Digitizing energy systems is not merely a tool replacement but a strategic transformation that turns energy data into critical business decision assets.
Designing an IoT-Based EMIS using Linear Regression and CUSUM for Real-Time Anomaly Detection in Pharmaceutical Industry Freska Lionia Darlion; Muhammad Isa Lufti
Jurnal PASTI (Penelitian dan Aplikasi Sistem dan Teknik Industri) Vol. 19 No. 3 (2025): Jurnal PASTI
Publisher : Universitas Mercu Buana

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22441/0.22441/10.22441/pasti.2025.v19i3.011

Abstract

The pharmaceutical industry is a high energy-intensity sector requiring strict operational stability. However, many companies still rely on manual energy monitoring methods, leading to information latency. A case study at a multinational pharmaceutical company in Indonesia revealed a baseload inefficiency of 14.3 kW during non-operational hours, which remained undetected due to an ad-hoc energy management system. This study aims to design an IoT-based Energy Management Information System (EMIS) architecture to transform the energy management business process from reactive to proactive-predictive. The study utilizes secondary data from the 2025 Energy Audit Report. The system design integrates a linear regression model (R²=0.80) for Energy Performance Indicators (EnPI) determination and the Cumulative Sum (CUSUM) algorithm for real-time anomaly detection. Investment feasibility is evaluated using techno-economic analysis. The implementation of EMIS requires an investment of IDR 225,000,000 with potential annual energy cost savings of IDR 44,172,687. Although the Simple Payback Period (SPP) is 5.1 years, the project is considered feasible due to its strategic value in data transparency, operational risk mitigation, and ISO 50001 compliance. Furthermore, this digital transformation supports the achievement of Sustainable Development Goals (SDG 7, 9, and 12) by promoting energy efficiency and responsible industrial consumption. Digitizing energy systems is not merely a tool replacement but a strategic transformation that turns energy data into critical business decision assets.