FG Automatisierungstechnik
Refine
Year of publication
Document Type
- Scientific journal article peer-reviewed (19) (remove)
Way of publication
- Open Access (1)
Keywords
- Asset-to-Go (1)
- Machine Learning (1)
- Mensch-Roboter-Kooperation (1)
- Ramp-up, Environmental impact, Knowledge management (1)
- Ramp-up; Knowledge management; Knowledge manager (1)
- Wandlungsfähigkeit (1)
- classification (1)
- feature selection (1)
- homogeneous feature selection ensemble (1)
- knowledge acquisition system (1)
Institute
- FG Automatisierungstechnik (19)
- FG Systeme (1)
Feature selection (FS) represents an essential step for many machine learning-based predictive maintenance (PdM) applications, including various industrial processes, components, and monitoring tasks. The selected features not only serve as inputs to the learning models but also can influence further decisions and analysis, e.g., sensor selection and understandability of the PdM system. Hence, before deploying the PdM system, it is crucial to examine the reproducibility and robustness of the selected features under variations in the input data. This is particularly critical for real-world datasets with a low sample-to-dimension ratio (SDR). However, to the best of our knowledge, stability of the FS methods under data variations has not been considered yet in the field of PdM. This paper addresses this issue with an application to tool condition monitoring in milling, where classifiers based on support vector machines and random forest were employed. We used a five-fold cross-validation to evaluate three popular filter-based FS methods, namely Fisher score, minimum redundancy maximum relevance (mRMR), and ReliefF, in terms of both stability and macro-F1. Further, for each method, we investigated the impact of the homogeneous FS ensemble on both performance indicators. To gain broad insights, we used four (2:2) milling datasets obtained from our experiments and NASA’s repository, which differ in the operating conditions, sensors, SDR, number of classes, etc. For each dataset, the study was conducted for two individual sensors and their fusion. Among the conclusions: (1) Different FS methods can yield comparable macro-F1 yet considerably different FS stability values. (2) Fisher score (single and/or ensemble) is superior in most of the cases. (3) mRMR’s stability is overall the lowest, the most variable over different settings (e.g., sensor(s), subset cardinality), and the one that benefits the most from the ensemble.
A maturity assessment approach for conceiving context-specific roadmaps in the Industry 4.0 era
(2019)
The digital transformation of the production sector is setting the scene for a major industrial change. The need for supporting companies in this transformation is currently covered by several maturity models, generally operationalized through standardized questionnaires, which provide, as an outcome, an assessment of the current maturity stage and a set of general improvement recommendations according to it. However, to provide companies with a more tangible support, there is a need for more individual approach. In order to deal with this need, this paper proposes, following a design science research framework, a novel approach based on Problem-Based Learning for structuring the assessment procedure as a dialectic process. This approach aims at facilitating the contextualization of the assessed company and, consequently, the identification of context-specific improvement recommendations. The proposed approach, supported by a maturity model used for framing information collected during the assessment process, is tested in three industrial cases. Although these have been assessed at the same maturity stage, different improvement recommendations have been proposed according to contextual factors such as strategic goals, core processes and key performance indicators.
Die heutige Automobilindustrie ist durch zunehmende Individualisierungswünsche der Kunden sowie einer Verkürzung der Konjunkturzyklen gekennzeichnet. Für Fahrzeughersteller, wie der Daimler AG äußert sich diese Marktsituation in einer hohen Variantenvielfalt und Stückzahlschwankungen. Da die Ausprägungen dieser Einflüsse weitgehend unbekannt sind, muss das Produktionssystem darauf vorbereitet sein, bedarfsgerecht der Marktsituation angepasst zu werden. Diese Anpassungen müssen schnell, mit wenig Aufwand und geringen Kosten möglich sein. Wandlungsfähige Produktionssysteme gehören deshalb zu den strategischen Erfolgsgrößen für die Automobilindustrie. Innerhalb dieses Beitrags wird ein modularer, skalierbarer Produktionsbaukasten vorgestellt, der Bestandteil einer Planungs- und Entwicklungsumgebung wandlungsfähiger roduktionskonzepte ist. Die hierfür genutzte grenzwerttheoretische Entwicklungsmethodik wird erläutert und die Skalierungsmöglichkeiten des Produktionsbaukastens mit Hilfe neuer Produktionstechnologien betrachtet. Dabei zeichnet sich die Mensch-Roboter-Kooperation als wesentlicher Bestandteil wandlungsfähiger Produktionskonzepte ab. „Smarte Produktionsassistenten“ sind intelligente und sichere Roboter-Applikationen, die sich als Möglichkeit hervortun die Mensch-Roboter-Kooperation als Schlüsseltechnologie in industriellen Anwendungen zu realisieren.
Nutzung von Erfahrungswissen im Fahrzeuganlaufmanagement. Bewertung und Auswahl geeigneter Methoden
(2013)
Elicitation of Requirements for a Knowledge Management Concept in Decentralized Production Planning
(2013)
The planning in manufacturing system is becoming
complicated day by day due to the expanding networks and shortage
of skilled people to manage change. Consequently, faster lead time
and rising demands for eco-efficient evaluation of manufacturing
products and processes need exploitation of new and intelligent
knowledge management concepts for manufacturing planning. This
paper highlights motivation for incorporation of new features in the
manufacturing planning system. Furthermore, it elaborates
requirements for the development of intelligent knowledge
management concept to support planning related decisions.
Afterwards, the derived concept is presented in this paper considering
two case studies. The first case study is concerned with the
automotive ramp-up planning. The second case study specifies
requirements for knowledge management system to support decisions
in eco-efficient evaluation of manufacturing products and processes.
The automotive industry is experiencing several limitations due to high mass customization on product side and the consequent evolution of newer ones as well as the expansion of existing production setups. There exist various complexities in planning and control of distributed production setups. This paper addresses issues related to planning using knowledge management approaches in the decentralized manufacturing domain with two distinct cases. The first case relates to the higher level planning methodology to automate experts’ knowledge in production ramp-up whereas the second case relates to optimization of production processes using knowledgebase as an intelligent assistance system for optimization.
This paper focuses on the systematic methodology for
incorporating intelligence and development methodology for
knowledge acquisition system in an automated manufacturing
environment. The intelligence is incorporated in the shape of
technology data catalogue that contains the knowledge about
production system as a whole. The knowledge acquisition system is
implemented in the form of a multiuser scalable interface into remote
human machine interface devices (e.g. Personal Digital Assistants)
with a purpose of extracting concrete and precise information and
knowledge about manufacturing systems and processes in highly
automated manufacturing environment. The extraction of precise
knowledge as well as organized access to the knowledge will
facilitates the operators, technicians and engineers for making faster,
safer and simpler on-process modifications and parameters
optimization.