Conduct interviews with more Field Service Engineers, especially with engineers in Japan consid-ering the high percentage of returns in that country. The population of Field Service Engineers should be a mix between engineers that return many parts and engineers that return less parts.
Gain more insights into the transportation costs. Currently only total transportation costs are known and not costs per order. Based on the interviews conducted it was observed that the numbers mentioned in this research are probably on the low end of the spectrum.
Gain more insights into the holding costs and order costs. These costs are important for inventory calculations. First, the proposed inventory control model would have better results if these costs could also be used as input parameters. Second, Servigistics would also profit from improved data.
Use the proposed control model to determine appropriate target fill rates and to re-evaluate the current stock levels. The output of the model indicated that some parts are stocked excessively and some parts insufficiently. In addition, too much stock is held in NA. This stock can be re-allocated to other warehouses.
As the values of the input parameters are important for the allocation of stock to parts, it is important that the company conducts research on input parameters and improves the accuracy of data in the ERP system. It was observed that the return lead times have not been updated for a long time. As shown in the sensitivity analysis, the return lead times are of great importance for the distribution of stock between the parts and warehouses. This should not only be done for XYZ tools, but also for the legacy tools.
Reduce the return lead times. Reducing return lead times would lead to lower inventory values.
However, if the company decides not to change the return lead times, then at least an effort should be made to make sure parts are returned in time.
Conduct research on the transportation lead times between the warehouses. Shorter transportation lead times between warehouses means that less inventory needs to be stocked in the inbetween hubs and local warehouses. For this to be re-evaluated it is also important that more insight is gained into the transportation costs.
Recommendation for future research in the company
Research the effect of incorporating part criticality in the model
Both the proposed inventory control model and the current software tool used to optimize stock
levels for the legacy tools use the aggregate fill rate per region to calculate the optimal stock levels.
As shown in the sensitivity analysis, this can lead to some parts having low fill rates, while other parts having high fill rates. This way of working does not distinguish between part criticality. It can therefore mean that high fill rates are possibly obtained for parts which are not necessarily critical. Incorporating criticality in the model can therefore be beneficial for the company.
Research the accuracy of the demand distribution.
If a significant number of orders consists of more than 1 part (i.e. random demand sizes), a compound Poisson distribution might provide a better fit to the data. If a compound Poisson distribution fits the data better, this would have implications for the inventory control model, since the proposed model assumes Poisson distribution. If a compound Poisson distribution has a better fit to the data, this would mean that the proposed control model overestimates fill rates.
The analysis could be done for both XYZ parts and legacy tools.
Research if lateral transshipments would lead to lower net stock.
Lateral transshipments refers to the situation where the regional and local warehouses are not only resupplied by the central warehouse, but also by other warehouses. Although infrastructure and compliance rules can complicate this process and reduce the benefits, it might be worth investigating. Especially for the APR region, where many smaller warehouses are situated at relatively close proximity.
References
Axs¨ ater, S. (2015). Inventory control (Vol. 225). Springer.
Basten, R. J. I., & van Houtum, G.-J. (2014). System-oriented inventory models for spare parts. Surveys in operations research and management science, 19 (1), 34–55.
de Kruijff, J. T. (2019). High-tech low-volume production planning (Ph.D. thesis). Eindhoven University of Technology, Eindhoven (The Netherlands).
Lu, L., & Yang, J. (2012). An inventory model for allocating repairable spares based on three-echelon supply relationship. In 2012 international conference on quality, reliability, risk, maintenance, and safety engineering (pp. 470–474).
Muckstadt, J. A. (2004). Analysis and algorithms for service parts supply chains. Springer Science & Business Media.
Ruan, M., Peng, Y., Li, Q., & ZHANG, G.-y. (2012). Optimization of three-echelon inventory project for equipment spare parts based on system support degree. Systems Engineering-Theory & Practice, 7 .
Rustenburg, J. W., van Houtum, G.-J., & Zijm, W. H. M. (2003). Exact and approximate analy-sis of multi-echelon, multi-indenture spare parts systems with commonality. In Stochastic modeling and optimization of manufacturing systems and supply chains (pp. 143–176).
Springer.
Sherbrooke, C. C. (1968). Metric: A multi-echelon technique for recoverable item control.
Operations Research, 16 (1), 122–141.
Thijssen, I. C. M. (2007). Multi-echelon spare parts management in europe at fei company (MSc thesis). Eindhoven University of Technology, Eindhoven (The Netherlands).
Topan, E., Bayındır, Z. P., & Tan, T. (2017). Heuristics for multi-item two-echelon spare parts inventory control subject to aggregate and individual service measures. European Journal of Operational Research, 256 (1), 126–138.
Van Strien, P. J. (1997). Towards a methodology of psychological practice: The regulative cycle. Theory & Psychology, 7 (5), 683–700.
Wong, H., Kranenburg, B., van Houtum, G.-J., & Cattrysse, D. (2007). Efficient heuristics for
two-echelon spare parts inventory systems with an aggregate mean waiting time constraint
per local warehouse. OR spectrum, 29 (4), 699–722.
Appendix
A Implementation of Model
The implementation of the model combined with the optimization from chapter 4 have been implemented in a decision support tool. The decision support tool enables the company to optimize stock levels for the central warehouse and two inbetween hubs and run different scenarios to see what happens when changing certain input values, such as return lead time, transportation lead time between central warehouse and inbetween hub etc. The tool has been built with PYQT5 in Python and a self-contained .exe has been created for Windows.
Input Parameters
Partnumber: The code of the part
Location name: The name of the warehouse. Note that the tool is designed for XYZ-parts and as explained in the thesis three warehouses are constructed: APR, NA and EU
Cost Of Goods Sold: The standard price of the part
Daily forecast: The daily forecast of the part
Procurement lead time : The procurement lead time of the part
Repair lead time: The repair lead time of the part
Economic Order Quantity: The EOQ of the part
Return lead time: The Return lead time of the part
Repairable: Whether the part is repairable or not
Output of the tool
Aggregate fill rate
Fill rate per part
Optimal stock level per part
Expected inventory on hand per part
Expected inventory value per part
Tool
The tool is opened by clicking on the .exe file (see figure A.1). Under ’File’ the user can select the Excel file with the input parameters. After selecting the Excel file the window in figure A.2 is displayed. Part numbers have been blurred out for confidentiality reasons.
Figure A.1: Screenshot when opening the tool
Under ’Input Data’ the values of the input parameter can be seen. The input parameters are displayed per warehouse. The user has the possibility to slide trough the warehouses and see parts per warehouse.
’Model config’ shows the target fill rates per warehouse. These can be adapted by simply typing the desired fill rate instead of the default value of ’0.9’.
The user can change the input parameters under ’Parameter Modifications’. Return lead time, procure-ment lead time, shipprocure-ment lead time between warehouses (default value is 14) and repair lead time can be adapted.
Additionally the user has the possibility to import stock levels (i.e. not let the tool calculate optimal base stock levels and reorder points, but import stock levels to see what the fill rate would be when these stock levels are used). This option can also be found under ’File’.
When the user wishes to run the program, he/she should go to ’Execute’. Under ’Execute’ the user can press ’Run’.
When the program is finished running, several Excel files are created in the same directory as where the tool is situated. These Excel files contain the output parameters described before.
An extensive work instruction has been provided to the company.