In English

Forecasting of Spare Parts Based on Vehicle Condition Monitoring Data A Case Study at Volvo Group

Anton Ottosson ; Olof Wireklint
Göteborg : Chalmers tekniska högskola, 2018. 75 s. Master thesis. E - Department of Technology Management and Economics, Chalmers University of Technology, Göteborg, Sweden; E2018:010, 2018.
[Examensarbete på avancerad nivå]

An accurate aftermarket demand forecast is critical for companies within the automotive industry. However, due to erratic and intermittent demand patterns of spare parts, achieving an accurate demand forecast through currently used time-based methods is difficult. Therefore, this thesis aims to predict future demand with causal-based forecasting methods using condition monitoring data, specifically fault codes, as explanatory variable. Furthermore, an evaluation is made on what effects an implementation would result in for the case company Volvo Group. The thesis method combines qualitative as well as quantitative data collection and analysis. Findings from the qualitative part of the study are a list of fault codes as well as a list of part types where a forecast based on condition monitoring data could be appropriate. To maximize potential economic benefits of an improved forecast the initial list of part types was subsequently filtered, which results in the part types: turbos and batteries. Those results were then validated through quantitative analysis using correlation and regression to determine causality between fault codes and spare part demand for three individual turbochargers. The causal relationships were used to create causal-based forecasts on different time horizons. These causalbased models were then compared with currently used time-based models, which show an overall better performance of the causal-based models. Based on the research findings it is discussed that the developed causal-based forecasting method is appropriate for parts with a positive demand trend. Furthermore, it is argued that the developed model is unable to accurately detect single period fluctuations since they are caused by customer behavior which can not be explained by fault codes. The thesis contributions include an analysis approach which can be used in future research, and recommendations for future actions within the case company to further develop the research findings.

Nyckelord: forecasting, spare part logistics, condition monitoring data, fault codes, correlation analysis, regression analysis

Publikationen registrerades 2018-06-08. Den ändrades senast 2018-06-08

CPL ID: 255060

Detta är en tjänst från Chalmers bibliotek