Optimize the Time a Mercedes-Benz Spends on the Test Bench Using Stacking Ensemble Learning

The automotive industry relies heavily on test bench evaluations to ensure product quality and compliance with safety standards. A test bench is a platform or setup where various components or systems of a vehicle are tested under controlled conditions. A Mercedes-Benz vehicle spends lot of time on the test bench and a lot of features impact the testing time. By utilizing anonymized data gathered from various tests and simulations, this study seeks to identify key features that significantly impact the testing duration. Therefore, this paper aims to optimize the duration a Mercedes- Benz vehicle spends on the test bench by leveraging anonymized features. To achieve this, machine learning algorithms will be employed to analyze and interpret these features, ultimately enabling the development of predictive models. In specific, a stacking ensemble model with Random Forest Regressor, Extra-Tree Regressor and Lasso Regressor is proposed to streamline testing processes, enhance efficiency and productivity in Mercedes- Benz manufacturing plants. The proposed stacking model is compared with individual machine learning models and it is observed that stacking model outperforms individual models attaining highest $R^{2}$ score of 0.4367, indicating its effectiveness in combining the strengths of multiple algorithms.