AutoML Approach for Decision Making in a Manufacturing Context

Numerous human- and/or machine actors interact in the value chain during a product’s life cycle from its design to its final removal. This implies that operational usage data of the product (seen as a system) must be made available to all concerned stakeholders through efficient informational chains. To face large amounts of data, Artificial Intelligence (AI) and especially Machine Learning techniques aim to assist stakeholders in their decision-making process. However, the latter are not necessarily experts in Machine Learning technics and have to rely on ML experts and data scientists to provide analytical assistance. To overcome this problem, a new approach called Automated Machine Learning (AutoML) has recently been proposed to save time and increase efficiency by automating traditional Machine Learning steps. This paper aims to provide decision assistance models for stakeholders, based on AutoML techniques. The proposed model is illustrated by a use case in the context of the automotive industry and especially in the design phase of cars.