Refine
Document Type
- Journal article (10)
- Conference proceeding (9)
- Book chapter (4)
- Working Paper (1)
Is part of the Bibliography
- yes (24)
Institute
- ESB Business School (16)
- NXT Nachhaltigkeit und Technologie (8)
- Technik (1)
Publisher
This study explores the application of the PatchCore algorithm for anomaly classification in hobbing tools, an area of keen interest in industrial artificial intelligence application. Despite utilizing limited training images, the algorithm demonstrates capability in recognizing a variety of anomalies, promising to reduce the time-intensive labeling process traditionally undertaken by domain experts. The algorithm demonstrated an accuracy of 92%, precision of 84%, recall of 100%, and a balanced F1 score of 91%, showcasing its proficiency in identifying anomalies. However, the investigation also highlights that while the algorithm effectively identifies anomalies, it doesn't primarily recognize domain-specific wear issues. Thus, the presented approach is used only for pre-classification, with domain experts subsequently segmenting the images indicating significant wear. The intention is to employ a supervised learning procedure to identify actual wear. This premise will be further investigated in future research studies.
The maintenance of special tools is an expensive business. Either manual inspection by an expert costs valuable resources, or the loss of a tool due to irreparable wear is associated with high replacement costs, while reconditioning requires only a fraction. In order to avoid higher costs and drive forward the automation process in production, a German gear manufacturer wants to create an automatic evaluation of skiving gears. As a sub-step of this automated condition detection, it is necessary for wheels to be automatically aligned within a vision-based inspection cell. In extension to a study conducted last year, further image preprocessing steps are implemented in this publication and a new alignment algorithm from the autoencoder family is evaluated. By using an additional synthetic dataset, previous limitations could be clarified. The results show that thorough data preparation is beneficial for all solution approaches and that neural networks can even beat a brute force algorithm.
In recent years, both fields, AI and VRE, have received increasing attention in scientific research. Thus, this article’s purpose is to investigate the potential of DL-based applications on VRE and as such provide an introduction to and structured overview of the field. First, we conduct a systematic literature review of the application of Artificial Intelligence (AI), especially Deep Learning (DL), on the integration of Variable Renewable Energy (VRE). Subsequently, we provide a comprehensive overview of specific DL-based solution approaches and evaluate their applicability, including a survey of the most applied and best suited DL architectures. We identify ten DL-based approaches to support the integration of VRE in modern power systems. We find (I) solar PV and wind power generation forecasting, (II) system scheduling and grid management, and (III) intelligent condition monitoring as three high potential application areas.
Forecasting intermittent and lumpy demand is challenging. Demand occurs only sporadically and, when it does, it can vary considerably. Forecast errors are costly, resulting in obsolescent stock or unmet demand. Methods from statistics, machine learning and deep learning have been used to predict such demand patterns. Traditional accuracy metrics are often employed to evaluate the forecasts, however these come with major drawbacks such as not taking horizontal and vertical shifts over the forecasting horizon into account, or indeed stock-keeping or opportunity costs. This results in a disadvantageous selection of methods in the context of intermittent and lumpy demand forecasts. In our study, we compare methods from statistics, machine learning and deep learning by applying a novel metric called Stock-keeping-oriented Prediction Error Costs (SPEC), which overcomes the drawbacks associated with traditional metrics. Taking the SPEC metric into account, the Croston algorithm achieves the best result, just ahead of a Long Short-Term Memory Neural Network.
Thorough maintenance of industrial equipment is crucial for the finances of companies. Whereas the purchase of new tools can be an expensive business, reconditioning special gear often costs just a fraction. In this paper, preliminary steps for an accurate visual based preventive maintenance of hobbing wheels are investigated. To perform robust and reliable decisions about the wheel's condition, tool department specialists require precise taken captures of it. For this reason, a visual control cell is built, which depends on correctly aligned hobbing wheels in its image acquisition construction. The tool needs to be placed on a turn-table and rotated, so that a single tooth is centered in the field-of-view of the camera mounted on a robot arm. For this alignment task, three different main approaches with various preprocessing steps are investigated, a brute-force algorithm, an orb-feature approach and an image regression model. The results show that even a brute-force algorithm can be outperformed by a moderate deep neural network.
Mithilfe von Modellen der erklärbaren Künstlichen Intelligenz (XAI) können die Rüstzeiten in der CNC-Werkzeugherstellung reduziert werden, wodurch nicht nur die betriebliche Effizienz und Produktionskapazität gesteigert, sondern auch die wirtschaftliche Rentabilität von KI-Projekten nachgewiesen werden kann.
Demand forecasting intermittent time series is a challenging business problem. Companies have difficulties in forecasting this particular form of demand pattern. On the one hand, it is characterized by many non-demand periods and therefore classical statistical forecasting algorithms, such as ARIMA, only work to a limited extent. On the other hand, companies often cannot meet the requirements for good forecasting models, such as providing sufficient training data. The recent major advances of artificial intelligence in applications are largely based on transfer learning. In this paper, we investigate whether this method, originating from computer vision, can improve the forecasting quality of intermittent demand time series using deep learning models. Our empirical results show that, in total, transfer learning can reduce the mean square error by 65 percent. We also show that especially short (65 percent reduction) and medium long (91 percent reduction) time series benefit from this approach.
The general conclusion of climate change studies is the necessity of eliminating net CO2 emissions in general and from the electric power systems in particular by 2050. The share of renewable energy is increasing worldwide, but due to the intermittent nature of wind and solar power, a lack of system flexibility is already hampering the further integration of renewable energy in some countries. In this study, we analyze if and how combinations of carbon pricing and power-to-gas (PtG) generation in the form of green power-to-hydrogen followed by methanation (which we refer to as PtG throughout) using captured CO2 emissions can provide transitions to deep decarbonization of energy systems. To this end, we focus on the economics of deep decarbonization of the European electricity system with the help of an energy system model. In different scenario analyses, we find that a CO2 price of 160 €/t (by 2050) is on its own not sufficient to decarbonize the electricity sector, but that a CO2 price path of 125 (by 2040) up to 160 €/t (by 2050), combined with PtG technologies, can lead to an economically feasible decarbonization of the European electricity system by 2050. These results are robust to higher than anticipated PtG costs.
Mit der Energiewende hat die Bundesregierung den Umbau der Energieversorgung begonnen. Da das Gelingen der Energiewende für die Zukunfts- und Wettbewerbsfähigkeit des Wirtschaftsstandorts Deutschland essenziell ist, wurden seitens des Bundesverbandes der deutschen Industrie (BDI) bereits 2013 Impulse für eine smarte Energiewende veröffentlicht, in denen fünf Prinzipien abgeleitet werden, die einen Rahmen für den Diskurs über die zu ergreifenden Maßnahmen setzen. Erneuerbare Energien werden in dem kommenden Jahren die dominierende Stromquelle darstellen. Daraus entstehen neue Herausforderungen. Zu deren Bewältigung hat das Bundeswirtschaftsministerium (BMWi) kürzlich eine 10-Punkte-Agenda (ZPA) für die zentralen Vorhaben der Energiewende vorgelegt. Zu diskutieren ist, inwieweit sie im Einklang mit den fünf Prinzipien des BDI steht und an welchen Stellen Anpassungen notwendig werden, damit der Umbau des Energiesystems erfolgreich gelingen kann.
The promise of the EVs is twofold. First, rejuvenating a transport sector that still heavily depends on fossil fuels and second, integrating intermittent renewable energies into the power mix. However, it is still not clear how electricity networks will cope with the predicted increase in EVs and their charging demand, especially in combination with conventional energy demand. This paper proposes a methodology which allows to predict the impact of EV charging behavior on the electricity grid. Moreover, this model simulates the driving and charging behavior of heterogeneous EV drivers which differ in their mobility pattern, decision-making heuristics and charging strategies. The simulations show that uncoordinated charging results in charging load clustering. In contrast, decentralized coordination allows to fill the valleys of the conventional load curve and to integrate EVs without the need of a costly expansion of the electricity grid.
