This paper presents a novel approach for detecting anomalies or faults in household appliances based on sensor data collected from a smart home environment. The proposed method leverages machine-learning techniques to analyze energy consumption patterns, environmental conditions, and weather data to identify abnormal appliance behavior. In this paper, five different anomaly detection algorithms have been used and the results are compared, analyzing their performance. The training and testing of the machine learning models is done on “Appliances energy prediction dataset” from the UCI Machine Learning Repository. How machine learning can be embedded in modern smart home automation systems is also discussed in this paper. We have comparing different anomaly detection models which can provide valuable insights for the maintenance and optimization of smart home systems. Comparing different models can help in detecting, irregular patterns in the smart home systems.