EoL stockpiling optimized for spare parts supply
Outstanding after-sales service is the way to stand out in the consumer electronics sector. This is why the supply of spare parts is very important to Medion. The products are subject to special requirements in terms of spare parts supply: Short product life cycles, high-priced spare parts and unpredictable spare parts requirements characterize this business.
Medion AG is one of the very few companies with a comprehensive portfolio of consumer electronics and information technology products. In terms of product groups, the company focuses on three pillars:
- PC/Multimedia,
- Consumer and household electronics,
- Communication technology.
The company operates according to the “build-to-order” principle, i.e. the devices are only produced in the required quantities once orders for sales promotions have been received from retail partners. The advantage of this strategy is that no storage costs are incurred – a cost saving that the company passes on to its customers and therefore impresses with a particularly good price-performance ratio. It also ensures that the company’s products are always up to date. In addition to Germany, Medion also has a strong geographical presence in the entire eurozone, including Scandinavia and the UK, and very well-positioned and professional sales and service units in conjunction with its retail and cooperation partners. Now you might think that such a business is quite simple: create and distribute product ideas, manufacture and deliver and that’s it. However, this is not the case.
CUSTOMER SERVICE AS THE HIGHLIGHT of the range of services.
Medion provides a particularly important element of the overall service for retailers and manufacturers through its own after-sales service for the end consumer, so that the partners do not have to worry about this business. A 365-day hotline in the company’s own call center provides expert support for questions about the application, warranty processing and repeat orders. The manufacturer also carries out all necessary repairs and, if necessary, commissions an on-site service for special products within Germany, which visits the consumer to provide advice or carry out repairs. The company guarantees a high availability of spare parts in order to be able to satisfy the end customer with fast problem solving. This means that Medion cannot manage entirely without a warehouse, as the correct stocking of replacement components is of particular importance. The particular challenge is that the spare parts cannot be procured for the entire warranty period. Manufacturers of the parts cancel these well before the end product’s warranty period expires. Motherboards for PCs are often only available in identical configurations for a few months. And plastic housings for notebooks, for example, are often only used during production itself. The appropriate hinge, should it be defective during the warranty period, must therefore also be ordered in good time. After all, you have to ensure that you have enough spare parts in stock for the rest of the warranty period. At this point in time, however, there is hardly any information available on the demand for spare parts, as the products have not been in use at the customer for long.
This harbors two dangers:
1. if stocks are too low, spare parts run out and customer appliances can no longer be repaired.
2. if stock levels are too high, Medion will be left with high inventories at the end of the warranty period, which will significantly increase costs.
The question is therefore how this conflict of objectives can be resolved. As part of the investigations, the consumption series of the spare parts were analyzed and evaluated. Due to the very short consumption periods (<6 to 12 months) and the very erratic demand, traditional methods such as averaging, exponential smoothing and median were not suitable for forecasting or did not offer any advantages over the method used to date. For this reason, past projects were examined more closely and the consumption series were analyzed with regard to their failure behavior. Special types of failure patterns emerged: A total of five different types were identified and assigned to so-called norm curves. Depending on the failure rates at the start of product use, in the mid-range and at the end of the warranty period, the components could be assigned to the standard curves.
TAKE A CLOSE LOOK.
It was no surprise that failure patterns were not necessarily specific to a particular assembly. Electronic components, for example, were to be found in all areas, as were mechanical components. The application of the standard curves could therefore not simply be regulated via the product group. Separate rules and indicators had to be used for this.
In the second step, methods were developed for using these standard curves to forecast spare parts requirements. As this was a very complex and computationally intensive task, a prototype of an analysis software tool was specially developed to process the failure data and determine forecasts based on the standard curves. By assessing the first phase of the consumption period, users can assign the most suitable failure pattern to the spare part and determine the requirement within the warranty period. The user also has options for manipulating the forecast. This makes it possible to map the user’s extensive experience via the tool. The forecast values are based on the standard curve. This can then be stretched or compressed with simple clicks or the consumption periods can be reweighted. The tool is very graphically oriented and provides very quick information about the consumption and demand situation.
In the final step, the methodology was validated. Forecasts were determined at certain points in the past on the basis of the information available at that time and compared with the forecasts made using the method applied to date. The result was clear: The new methodology now makes it possible to make better predictions about spare parts requirements. In particular, it was possible to improve the coverage of residual requirements by a double-digit percentage, which on the one hand noticeably improves delivery readiness and at the same time reduces inventories, making this project a double success. In the next step, this new methodology will be implemented in the standard product Diskover SCO for forecasting and scheduling optimization, so that corresponding functionalities can already be used for spare parts management with the standard software.
