Second work report from Use Case 2: “Data-driven networked quality management”

Objective

Use Case 2, “Data-driven, networked quality management”, aims to lay the foundations for a gradually expandable, efficient and holistic reporting and analysis system. This system will create transparency about the quality situation and focus on potentials for quality improvement and optimization of products and processes. This will not only be done in a way that is specifically differentiated to the users, but will also be developed with and by them.

Development of generalized analysis modules

In the previous first work report, the development of analysis modules for alarm and forecast was presented using a specific example (spare parts sales) as well as the underlying methodology. These two typical analysis questions in the standardized problem-solving process and are at the center of the focus topic Q2025 (see figure 1), in which users at Miele are developing their own analysis modules with the support of the application partner RapidMiner. In the next step, a further set of alarm and forecast modules was developed, which should be much more general and thus more universally applicable. The same methodology was used for the development and evaluation as in the first case. The application of the building blocks and their value for different user groups was evaluated for different levels of customer service data from international sources. The result is very positive.

Figure 1: Miele focus topic Q2025

In addition, work carried out by the research partner IPS is progressing in a separate focus topic on networked devices. Here, different approaches to the formation of key figures are being developed and made available in AKKORD. Based on the data of the networked devices, a framework was developed that examines the determination of key figures as a time series. In this case, the topic of the networked devices is a good example of how the time series of different key figures result in different types of progression. Additionaly, to the classic, characteristic types of progression (trend-like, stationary, etc.), mixed forms also occur here. In the context of this field of work, therefore, different indicator curves are examined in order to select and develop appropriate types of monitoring. This work is being carried out specifically on the basis of questions on networked devices from the Business Unit Laundry, but due to the AKKORD approach of modularity, this general investigation of the processes with regard to practical application in quality management is also relevant for other areas at Miele as well as beyond the company. At the same time, the developed modules serve as pilot modules for handling in the AI Toolbox of the AKKORD platform. The next phase will serve to validate the key figures and to further develop the modules, especially in the direction of forecasting.

Rollout strategies: Benefit video

Systematic support (“change management”) of these changes is helpful so that the results and findings can be sustainably anchored in Miele’s quality management. One component of this is a benefit video, which is being developed together with the application partner mosaiic. It is intended to improve the visibility and perception of the use case, both within the Miele company and on the AKKORD SharePoint. On the one hand, it serves to communicate the AKKORD platform, and on the other hand, it helps to make interested employees at Miele aware of the importance of Citizen Data Science in quality management and beyond.

To explain: Experts from different areas have always worked together on interdisciplinary issues to ensure Miele quality. To this end, they analyze large quantities of data from “field observation” – always taking data protection into account, of course. This data provides concrete information from the usage phase of the appliances. It comes, for example, from customer service calls, from the networked appliances themselves and from Miele’s elaborate service life tests. With increasing availability and growing volume of different data, the potential added value but also the complexity of dealing with this data is increasing and will continue to do so in the future. Professionals in corporate divisions who build up the corresponding competencies are best able to tap into this added value. They are thus growing into a new role, also called Citizen Data Scientist.

A nice side effect: The Miele use case already shows how personal competences can be built up among employees with the help of piecework. Not only in quality management, but across all departments

Outlook

The collaboration in the set-up phase of the AKKORD platform has already resulted in a variety of analytical methods that can be used in data-driven networked quality management. In the remaining term of the project (until the end of March 2023) these different ways will be tested holistically. Through the benefits of video and the meanwhile advanced level of maturity of the platform, the circle of users should expand significantly. The users will then apply the analysis modules that will then be available in business practice, or even create new ones. Along the way, they will continue their education on the work & learn platform and thus contribute to the validation of the modular system.

Author and contact person:

Share This Story, Choose Your Platform!