Outlier Detection Approach in Sensor-to-Microcontroller Interfaces

This paper presents the implementation of an outlier detection method based on the Grubbs test, specifically designed for microcontroller applications. The theoretical background for outlier detection is outlined, and experimental validation is conducted using a direct sensor-to-microcontroller interface circuit. Results demonstrate that this approach improves the accuracy of measurement systems affected by Gaussian noise or other normal-distributed external interference signals.