Development of a Self-X Sensory Electronics for Anomaly Detection and its Conceptual Implementation on an XMR-based Angular Decoder Prototype

Abstract

Efficient fault diagnosis, early anomaly detection, and prevention of equipment failures are crucial for reducing production costs and minimizing unplanned downtime. This paper focuses on developing a self-X system using artificial immune system algorithms to address the challenges of condition monitoring, quality control, and predictive maintenance in Industry 4.0. The proposed methodology incorporates self-monitoring, self-healing, and self-repairing capabilities into intelligent measuring systems equipped with a tunnel magnetoresistive (TMR) sensor-based angular decoder. In particular, the extension of the self-X hierarchy to the adaptive electronics level, with a focus on healing/adjustments before ADC's irreversible quantization loss is a key objective. The implemented self-X approach enables dynamic offset and gain compensation, improving angle accuracy. The experimental setup involves a reconfigurable analog front end with self-X properties (AFEX), a data acquisition unit, feature extraction, self-monitoring, and self-healing mechanisms. The results demonstrate the successful implementation of gain compensation using a fabricated current-feedback instrumentation amplifier. The experiments show the impact of signal amplitude drop on angle error calculation, with the maximum absolute error observed at 11mm TMR sensor position. The self-X loop, including electronics, reduced the angle error to approximately 50 % by increasing the gain from 4 to 32. The results obtained from first experiment indicate a maximum absolute error of -6.552 ° in angle computation, which is not yet competitive to SoA systems and needs further consideration and system fine-tuning to achieve a competitive and self-X sensor system.