Trustworthy data-driven prognostics in gas turbine engines are crucial for safety, cost-efficiency, and sustainability. Accurate predictions depend on data quality, model accuracy, uncertainty estimation, and practical implementation. This work discusses data quality attributes to build trust using anonymized real-world engine data, focusing on traceability, completeness, and representativeness. A significant challenge is handling missing data, which introduces bias and affects training and predictions. The study compares the accuracy of predictions using Exhaust Gas Temperature (EGT) margin, a key health indicator, by keeping missing values, using KNN-imputation, and employing a Generalized Additive Model (GAM). Preliminary results indicate that while KNN-imputation can be useful for identifying general trends, it may not be as effective for specific predictions compared to GAM, which considers the context of missing data. The choice of method depends on the study’s objective: broad trend forecasting or specific event prediction, each requiring different approaches to manage missing data.
From the article: Though organizations are increasingly aware that the huge amounts of digital data that are being generated, both inside and outside the organization, offer many opportunities for service innovation, realizing the promise of big data is often not straightforward. Organizations are faced with many challenges, such as regulatory requirements, data collection issues, data analysis issues, and even ideation. In practice, many approaches can be used to develop new datadriven services. In this paper we present a first step in defining a process for assembling data-driven service development methods and techniques that are tuned to the context in which the service is developed. Our approach is based on the situational method engineering approach, tuning it to the context of datadriven service development. Published in: Reinhartz-Berger I., Zdravkovic J., Gulden J., Schmidt R. (eds) Enterprise, Business-Process and Information Systems Modeling. BPMDS 2019, EMMSAD 2019. Lecture Notes in Business Information Processing, vol 352. Springer. The final authenticated version of this paper is available online at https://doi.org/10.1007/978-3-030-20618-5_11.
MULTIFILE
Due to societal developments, like the introduction of the ‘civil society’, policy stimulating longer living at home and the separation of housing and care, the housing situation of older citizens is a relevant and pressing issue for housing-, governance- and care organizations. The current situation of living with care already benefits from technological advancement. The wide application of technology especially in care homes brings the emergence of a new source of information that becomes invaluable in order to understand how the smart urban environment affects the health of older people. The goal of this proposal is to develop an approach for designing smart neighborhoods, in order to assist and engage older adults living there. This approach will be applied to a neighborhood in Aalst-Waalre which will be developed into a living lab. The research will involve: (1) Insight into social-spatial factors underlying a smart neighborhood; (2) Identifying governance and organizational context; (3) Identifying needs and preferences of the (future) inhabitant; (4) Matching needs & preferences to potential socio-techno-spatial solutions. A mixed methods approach fusing quantitative and qualitative methods towards understanding the impacts of smart environment will be investigated. After 12 months, employing several concepts of urban computing, such as pattern recognition and predictive modelling , using the focus groups from the different organizations as well as primary end-users, and exploring how physiological data can be embedded in data-driven strategies for the enhancement of active ageing in this neighborhood will result in design solutions and strategies for a more care-friendly neighborhood.
Production processes can be made ‘smarter’ by exploiting the data streams that are generated by the machines that are used in production. In particular these data streams can be mined to build a model of the production process as it was really executed – as opposed to how it was envisioned. This model can subsequently be analyzed and stress-tested to explore possible causes of production prob-lems and to analyze what-if scenarios, without disrupting the production process itself. It has been shown that such models can successfully be used to diagnose possible causes of production problems, including scrap products and machine defects. Ideally, they can even be used to model and analyze production processes that have not been implemented yet, based on data from existing production pro-cesses and techniques from artificial intelligence that can predict how the new process is likely to be-have in practice in terms of data that its machines generate. This is especially important in mass cus-tomization processes, where the process to create each product may be unique, and can only feasibly be tested using model- and data-driven techniques like the one proposed in this project. Against this background, the goal of this project is to develop a method and toolkit for mining, mod-elling and analyzing production processes, using the time series data that is generated by machines, to: (i) analyze the performance of an existing production process; (ii) diagnose causes of production prob-lems; and (iii) certify that a new – not yet implemented – production process leads to high-quality products. The method is developed by researching and combining techniques from the area of Artificial Intelli-gence with techniques from Operations Research. In particular, it uses: process mining to relate time series data to production processes; queueing networks to determine likely paths through the produc-tion processes and detect anomalies that may be the cause of production problems; and generative adversarial networks to generate likely future production scenarios and sample scenarios of production problems for diagnostic purposes. The techniques will be evaluated and adapted in implementations at the partners from industry, using a design science approach. In particular, implementations of the method are made for: explaining production problems; explaining machine defects; and certifying the correct operation of new production processes.
Met het project Circl-Wood willen projectpartners Fijnhout en Nijboer, samen met de Hogeschool van Amsterdam (HvA) kennis ontwikkelen over het ontwerpen en produceren van hoogwaardige objecten uit afvalhout (Dit kan afvalhout betreffen uit verschillende bronnen; afvalinzameling, woningrenovatie, recycling bedrijven, maar ook reststukken van houtleveranciers en houtverwerkende bedrijven) met behulp van geavanceerde numerieke ontwerpgereedschappen (“computational design”). De projectpartners willen samen onderzoeken of het mogelijk is om hoogstaande en in het oog springende circulaire objecten te ontwerpen van een specifieke hoeveelheid afvalhout, met de HvA ligstoel - die in 2018 in een eerder KIEM project is gemaakt - als iconisch voorbeeld. Hierbij worden de kenmerken van het beschikbare hout (kleur, vorm, nerfrichting, houtsoort) als ‘data’ gebruikt om met ontwerpalgoritmes objecten te ontwikkelen met unieke kenmerken. Deze data-gedreven ontwerpmethode dient toepasbaar te zijn op een willekeurige batch hout die door robots geïnventariseerd en gesorteerd is. Het automatiseren van het ontwerpproces voor hoogwaardige producten creëert nieuwe circulaire toepassingsmogelijkheden voor afvalhout. In 2018 was het ontwerp van de stoel niet gebaseerd op de specifieke stukken hout waarvan hij werd gemaakt. Pas na het ontwerp werden stukken afvalhout handmatig geselecteerd, op maat gezaagd en verbonden tot een omhullende vorm, die door de robot 3D gescand is en waar vervolgens door de robot de stoel uit gefreesd is. In Circl-Wood echter wordt een geavanceerd ontwerpproces ontwikkeld: de gegevens van een beschikbare hoeveelheid resthout worden gebruikt om verschillende specifieke ontwerpen te maken met kleurpatronen, vormen en structuren gerelateerd aan het beschikbare hout. Het doel is om haalbare ontwerpen te berekenen op basis van het beschikbare hout. Het project demonstreert hoe numerieke ontwerpgereedschappen bij kunnen dragen aan een creatieve en efficiënte benutting van resthout van houtverwerkende bedrijven zoals Fijnhout voor interieur toepassingen (door bedrijven als Nijboer).