Development of Hybrid AI model for Car Steering Shaft Assembly by Combining Gaussian Process Regression and Artificial Neural Network

This paper presents a case study to apply artificial intelligence (AI) to the assembly of automotive parts. The sliding load of a car steering shaft assembly is controlled by selecting an appropriate size of the ball slider corresponding to the shaft and the tube. The manual assembly currently conducted by skilled workers has a low selection accuracy and long process time, which is the bottleneck of the whole manufacturing process. To increase the selection accuracy, an expert system based on a hybrid AI model was developed by combining Gaussian process regression and artificial neural network. The AI-based system could recommend suitable ball size corresponding to the over ball diameters measured on the tube and shaft. The system achieved 91.32% prediction accuracy in the test cases.