This thesis investigates how sampling rates, cutting disk RPS, and blade condition affect accelerometermeasurements in a robotic mower. Fifty test cases were conducted in both indoor and outdoor envi-ronments to evaluate how mechanical vibrations and operating environments influence slow acceler-ation detection. The results show that overall IMU environmental disturbances had a stronger impacton measurement reliability than individual disturbance factors while increasing sampling rates did notconsistently improve performance. Time domain metrics, inverse RMSE and inverse MAE, providedthe most reliable evaluation, while FFT offered mainly qualitative insight. Overall, both metrics werefound suitable for assessing accelerometer reliability under different operating conditions.