Machine Learning-based Real-Time Sensor Drift Fault Detection using Raspberry Pi

From smart industries to smart cities, sensors in the modern world plays an important role by covering a large number of applications. However, sensors get faulty sometimes leading to serious outcomes in terms of safety, economic cost and reliability. This paper presents an analysis and comparison of the performances achieved by machine learning techniques for realtime drift fault detection in sensors using a low-computational power system, i.e., Raspberry Pi. The machine learning algorithms under observation include artificial neural network, support vector machine, naive Bayes classifier, k-nearest neighbors and decision tree classifier. The data was acquired for this research from digital relative temperature/humidity sensor (DHT22). Drift fault was injected in the normal data using Arduino Uno microcontroller. The statistical time-domain features were extracted from normal and faulty signals and pooled together in training data. Trained models were tested in an online manner, where the models were used to detect drift fault in the sensor output in real-time. The performance of algorithms was compared using precision, recall, f1-score, and total accuracy parameters. The results show that support vector machine (SVM) and artificial neural network (ANN) outperform among the given classifiers.