PREDICTIVE MAINTENANCE PLATFORM DESIGN USING IOT ON RAILWAY TURNOUT SYSTEM

• Main Author: Sutani
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/80808
• Institution: Institut Teknologi Bandung

The significant public interest in utilizing rail transportation necessitates careful consideration of the condition of railway facilities and infrastructure to uphold service quality for passengers. The government's initiative to enhance train speeds and services will inevitably impact the state of railroad facilities and infrastructure. This study focuses on the turnout system, crucial for guiding trains along their intended paths. Regular inspection of the turnout system is imperative to prevent derailments, collisions, schedule disruptions, and other issues that hinder railway performance. This research aims to develop a platform integrating planned maintenance procedures with real-time analysis using lidar and accelerometer sensors at the point machine, identifying anomalies, particularly in roller and motor components. Platform testing, conducted in a laboratory/workshop setting using point machine simulation equipment, yielded promising results: a throughput of 270,000 per second with a test sample of 3000 threads and a response time of 1742 responses per millisecond. These findings are deemed satisfactory and can serve as a benchmark for predictive maintenance efforts.