A Model for Robot Arm Pattern Identification using K-Means Clustering and Multi-Layer Perceptron

Predictive maintenance of industrial machines is one the challenging applications in Industry 4.0. This paper presents a comprehensive methodology to identify robot arm (SCARA) movement patterns detect mechanical aging robot, which determined by abnormal arm. The dataset used two movements that go from point A B and then back A. Accelerometer data measure signal SCARA actions, mainly focus on non-linear movement. identification pattern made combining k-means multilayer perceptron. proposed approach first extracts valuable features as characteristics datasets time domain statistical value parameters. K-means clustering technique initiated label training dataset. In this phase, elbow curve determine number clusters dataset, 2 clusters. Moreover, assumption cluster labeled normal Hence, perceptron predict testing model yields an accuracy 94.14%, whereas its cross-validation 96.12%.