Machine Learning Based Cost Engineering of Automotive Parts – Lessons Learned

Bodendorf F, Franke J (2019)

Publication Language: German

Publication Type: Conference contribution, Abstract of lecture

Publication year: 2019

Publisher: Springer Nature

City/Town: Switzerland

Pages Range: 60

Conference Proceedings Title: Digitale Welt 4, 60

Event location: Munich DE

DOI: 10.1007/s42354-019-0234-3

Open Access Link: https://link.springer.com/article/10.1007/s42354-019-0234-3

Abstract

Strategic change management is one of the crucial fields of cost analytics in the automotive industry. Here, one important task is to explore potentials and propose measures to reduce costs of parts in the product development process. This is usually done within cost engineering and controlling departments of car manufacturing companies. The costs of part modifications are analyzed and specified usually by complex calculation methods based on a lot of input parameters describing the properties of a specific part. The calculation processes are mostly executed bottom-up. This means that                                                          the cost of a part is determined by going through many attributes and assigning and accumulating the corresponding cost estimations. These attributes comprise features of CAD designs, built-in materials as well as production sequences.

That is why the traditional calculation processes are effortful and time consuming. In addition, they are predominantly carried out manually with the help of simple calculation software like spreadsheets. Actually, Machine Learning (ML) algorithms and models are occupying center stage to cope with this problem and make the process more efficient. The overall goal is to predict costs of parts in the early phase of the product development process. In this early phase the expenditure of part modifications is low, but detailed information on the structure of the part is scarce. Machine Learning models are expected to be able to learn from a large amount of historical data and predict costs without using bottom-up calculation formulas. Also, Machine Learning algorithms are expected to find out the most influencing technical and economic parameters and in this way are able to reduce the considered input data to factors that are at hand in the early stage of the development process.


Authors with CRIS profile

Frank Bodendorf Lehrstuhl für Fertigungsautomatisierung und Produktionssystematik (FAPS) Jörg Franke Lehrstuhl für Fertigungsautomatisierung und Produktionssystematik (FAPS)

How to cite

APA:

Bodendorf, F., & Franke, J. (2019, November). Machine Learning Based Cost Engineering of Automotive Parts – Lessons Learned. Paper presentation at First International Symposium on Applied Artificial Intelligence (ISAAI), Munich, DE.

MLA:

Bodendorf, Frank, and Jörg Franke. "Machine Learning Based Cost Engineering of Automotive Parts – Lessons Learned." Präsentiert bei First International Symposium on Applied Artificial Intelligence (ISAAI), Munich Hrg. Digitale Welt Convention, Switzerland: Springer Nature, 2019.

BibTeX: Download