Assessing product quality from the production process health status

Le Toan Duong (1, 2) , Louise Travé-Massuyès (1) , Audine Subias (1) , Nathalie Barbosa Roa (1, 2)

1 LAAS-DISCO - Équipe DIagnostic, Supervision et COnduite
2 Vitesco Technologies

In the manufacturing industry, it is of prime importance to be able to retrieve the health status of production lines and to know how products navigate across operations. Products that are suspected to be faulty are deviated from the nominal path and inspected more closely. The fact that some products deviate from the nominal path and others fail at some check operations might indicate a risk for quality. This paper proposes a method to obtain the process model automatically , following the principles of the process mining field. The found model not only shows the product transitions from one operation to another, but also includes the time constraints bounding the "normal" time for each transition. From this model, products can be clustered according to the path they follow along the process. The clustering results for the product batch and the product itself are ultimately used to compute a quality index. The whole method is applied to a SMT PCB production line from Vitesco Technologies.

Domaines: Sciences de l'ingénieur [physics]

Le Toan Duong, Louise Travé-Massuyès, Audine Subias, Nathalie Barbosa Roa. Assessing product quality from the production process health status. 31st International Workshop on Principles of Diagnosis: DX-2020, Sep 2020, Nashville, Tennessee, United States. ⟨hal-03011772⟩