2025 International Conference on Sensing, Measurement & Data Analytics in the era of Artificial Intelligence (ICSMD 2025)

Condition monitoring (CM) is vital for ensuring the reliability and safety of rotating machinery in industrial applications. Despite significant advances in signal processing and machine learning techniques, progress in CM research remains constrained by the scarcity of high-quality degradation datasets. Acquiring such data, particularly for fault prognostics, faces multiple challenges. First, constructing an experimental test rig requires substantial investment and multidisciplinary expertise in mechanical design, control systems, and measurement technologies. Second, generating meaningful prognostic datasets demands large numbers of bearing failures under diverse operating conditions to adequately represent different degradation patterns. Third, the degradation process often takes several days to months, making it both time- and resource-intensive. As a result, few comprehensive datasets are publicly available due to commercial confidentiality and limited test capacities. To address these challenges and contribute to the CM community, we present the KU Leuven Bearing Prognostics Setup (KUL-BPS), a novel, modular, and cost-effective Accelerated Lifetime Testing (ALT) platform designed for systematic bearing degradation testing. Unlike existing ALT setups, which are often limited to low loads (<10 kN), low speeds (<3000 RPM), or artificially induced faults, KUL-BPS offers several key advantages: (a) It supports high loads up to 35 kN and speeds up to 18,000 RPM, enabling realistic wear and fatigue processes; (b) It utilizes a robust self-aligning bearing, which is easy to install, inspect, and replace; (c) It allows experiments under both constant and time-varying load-speed profiles to reflect real-world operating conditions. Furthermore, KUL-BPS enables synchronized acquisition of multiple sensing modalities, including vibration, torque, speed, force, acoustics, displacement, and temperature, facilitating comprehensive CM studies. The platform also supports virtual sensing research, such as estimating internal forces or torques from indirect measurements.
 

Condition monitoring,Degradation data,Accelerated Lifetime Test,Rotating Machinery,Fault Prognostics,Remaining useful life (RUL)