Advanced Technologies for Efficient Asset Tracking and Management

In the context of IoT-based industrial systems, asset reliability is vital for efficient supply chain management. As more sensors are deployed on equipment and machinery, fault detection in raw sensor data has become a more highlighted concern. This chapter focuses on improving predictive maintenance by identifying anomalies in the sensor output that are usually symptomatic of impending faults. Depending on the application of signal processing methodologies to identify useful patterns in the data, and followed by comparative evaluation of data analytics methods, the research describes how some of the data analytics methods are better suited for anomaly detection. Machine learning methods such as XGBoost and Isolation Forest were found to be highly effective in the detection of anomalous behavior in sensor networks. These findings help improve the uptime and reliability of mission-critical assets, and therefore, more efficient and responsive supply chain operations in industrial economies.