Generalizing Vehicle Manoeuvre Prediction Across Diverse Datasets

Due to the complex nature of automotive components and sensor data, predictive maintenance is essential to ensure the reliability and safety of the vehicle. This work introduces a new predictive algorithm for automotive engine health, designed as it will provide higher accuracy and faster decisions in detecting potential engine failures, linear Through discriminant analysis, Gaussian naive edges, support vector machines, decision trees, random forests, gradient enhancement, and AdaBoost, the program displays patterns and abnormalities that may indicate impending engine problems.The data set undergoes extensive preprocessing steps such as standardization, handling missing values, and feature engineering to improve model performance. The evaluation criteria used include accuracy, precision, and confusion matrix, with special attention to prevent overfitting through regularization and the early stop method. In the developed model, the group method, especially stacked model 1, obtains impressive results with a model accuracy of 0.99. This high accuracy highlights the effectiveness of the ensemble approach in managing forecasts. The model’s ability to deliver real-time analysis and early warning can help significantly reduce maintenance costs, prevent failures, and enhance vehicle safety, resulting in improved vehicle engine health during the maintenance process.