A Microcontroller Operating Strategy for (Micro-)Pitting and Temperature Increase Detection in Sensor-Integrating Gears Evaluated with Pre-recorded Sensor Data

Condition-based gear health monitoring is crucial to prevent unexpected machine downtimes. However, the lack of small-scale and cost-effective sensors that can be easily integrated on a system level hinders the implementation of condition-monitoring approaches. Damage to gears often occurs due to tooth contact, making it essential to acquire accurate data close to the gear engagement for reliable damage detection. Integrating sensors directly into gears is a promising solution for a space-neutral sensor system, but continuous data acquisition is limited by the energy and memory constraints of Microcontrollers (MCUs). This paper proposes and evaluates an operating strategy for detecting gear damage and temperature increases, featuring automated gear state detection to reduce energy consumption. Testing on a Cortex-M0+ MCU demonstrates its suitability for lowpower devices. The strategy was tested using down-sampled pre-recorded acceleration data measured by external, highperformance sensors. An optimal MCU clock frequency is determined by evaluating its impact on energy consumption and execution time. The strategy’s energy consumption and execution time for different states are presented, highlighting opportunities for future optimization.