In industrial drive systems, induction motors and their position sensors, such as resolvers, are essential components that directly influence operational stability and productivity. Failures in these elements can lead to system downtime, product quali...
In industrial drive systems, induction motors and their position sensors, such as resolvers, are essential components that directly influence operational stability and productivity. Failures in these elements can lead to system downtime, product quality degradation, and increased maintenance costs. In particular, inter-turn short fault (ITSF) in induction motors are difficult to detect at an early stage because the resulting variations in current and vibration signals are extremely subtle, while resolvers are highly sensitive to structural and mechanical misalignments, where even slight eccentricity or runout can introduce significant position estimation errors, degrading control performance and system reliability. To address these industrial challenges, this study proposes two experimental, real-time diagnostic techniques that target electrical and mechanical failures in motors and sensors, respectively. First, to overcome the limitations of conventional MCSA- and vibration-based monitoring methods that struggle to identify early-stage ITSF, a flux-based diagnostic approach is introduced by directly measuring the axial stray flux of the motor. This method enables robust detection of early inter-turn short indications using only non-invasive external sensing. Second, to diagnose static eccentricity, dynamic eccentricity, and runout faults in VR resolvers without relying on frequency-domain analysis or complex signal processing, a Lissajous-curve-based geometric interpretation of the resolver’s sine/cosine output is proposed. This approach allows intuitive and consistent detection and classification of mechanical faults through simple geometric indices. The two techniques provide reliable assessments of electrical faults in motors and mechanical faults in resolvers using lightweight signal processing, and together they satisfy the industrial requirements of real-time operation, non-invasiveness, and robustness. The results of this study offer a practical foundation for implementing condition- based maintenance (CBM) in industrial drive systems by enabling integrated monitoring of both actuators and position sensors.