Riassunto analitico
Hybridization of transmission represents an attractive solution for HyperCars manufacturers in order to improve performances of their vehicles. The introduction of this technology increased the complexity of the systems, hence their reliability must be controlled. In particular, the End-of-Line test is an important step of the production line that allows car manufacturer to guarantee quality and performances of their products. In order to have a good control on this phase, the ideal procedure is to perform an accurate investigation before the beginning of production. However, in many business realities, the low volumes of production are not sufficient to justify this kind of investments. The aim of this project was to strengthen the process with the identification of any possible deficiency of the actual system. In order to achieve the target, the entire data set was analyzed. Firstly, the development of a software allowed to perform an analytic offline analysis in an automatic way. This approach allowed to discover some limits of the test bench and to identify some deficiencies of the products. Then, the increasingly interest in Artificial Intelligence of the last years, drove the project to the introduction of a further analysis with AI. This expedient was used to deeply detect correlations between events, highlighting the causes of unwanted behavior.
|
Abstract
Hybridization of transmission represents an attractive solution for HyperCars manufacturers in order to improve performances of their vehicles. The introduction of this technology increased the complexity of the systems, hence their reliability must be controlled. In particular, the End-of-Line test is an important step of the production line that allows car manufacturer to guarantee quality and performances of their products. In order to have a good control on this phase, the ideal procedure is to perform an accurate investigation before the beginning of production. However, in many business realities, the low volumes of production are not sufficient to justify this kind of investments. The aim of this project was to strengthen the process with the identification of any possible deficiency of the actual system. In order to achieve the target, the entire data set was analyzed. Firstly, the development of a software allowed to perform an analytic offline analysis in an automatic way. This approach allowed to discover some limits of the test bench and to identify some deficiencies of the products. Then, the increasingly interest in Artificial Intelligence of the last years, drove the project to the introduction of a further analysis with AI. This expedient was used to deeply detect correlations between events, highlighting the causes of unwanted behavior.
|
