Skip to content
printicon
Main menu hidden.

Statistical learning for an automatic alarm system and root cause analysis of defects when painting truck cabins

PhD project at the Industrial Doctoral School at Umeå University.

The goal of Niklas Fries' project is to develop statistical methods to explain and prevent quality problems when painting truck cabins.

Head of project

Niklas Fries
Doctoral student
E-mail
Email

Project overview

Project period:

2017-10-01 2022-09-30

Funding

The project is financed by Umeå University (50 percent) and Volvo Lastvagnar AB (50 percent).

Participating departments and units at Umeå University

Department of Mathematics and Mathematical Statistics

Research area

Mathematics, Statistics

Project description

In my research, we study the painting process in Volvo Trucks' cab factory in Umeå. For this process, I want to collect data that describe the process as well as data that describe the resulting quality outcome.

Based on these data, I will develop methods for constructing cabin-specific explanatory variables as well as response variables.

When I have these cabin-specific variables, I want to use them to identify which process parameters have the largest influence on the quality of the painting.

When this is done, I want to develop an automatic alarm system that identifies an increased risk of quality problems, and a system for root cause analysis that identifies probable causes and proposes corrective measures.