AI-Powered Predictive Maintenance System

A transportation organization in Japan needed to monitor motor operating conditions and catch anomalies across varying stress levels and workloads. We built an AI-based anomaly detection system that applies time-series analysis to three-axis vibration and workload sensor data, delivering automated alerts and continuous condition monitoring. Our system flags faults in under 100 ms, cuts manual inspection effort by 85%, and reduces inspections by 50%.

Challenge

• Needed to monitor motor operating conditions in real time across varying stress levels and workloads
• Required early detection of anomalies from high-frequency sensor data before failures occurred
• Manual inspection was labor-intensive and could not scale to continuous monitoring

Solution

• Built an AI-based motor anomaly detection system using vibration and workload sensor signals
• Applied time-series analysis to vibration and workload data across three axes (X, Y, Z)
• Processed frequency data from motors under multiple operating conditions
• Delivered automated anomaly alerts alongside continuous condition monitoring, improving detection reliability

Architecture

Outcome

• Under 100 ms detection latency for real-time anomaly alerts
• 85% reduction in manual inspection effort
• 50% fewer inspections through condition-based monitoring
• More consistent fault detection across varying operating conditions

Tech Stack

• Backend: Python
• Data: Sensor Data Processing
• AI/ML: TensorFlow, Time-series ML