Kählert, Alexander (2017)
Specification and Evaluation of Prediction Concepts in Aircraft Maintenance.
Technische Universität Darmstadt
Dissertation, Erstveröffentlichung
Kurzbeschreibung (Abstract)
The goal of this thesis is to identify and quantify the potentials of predictive maintenance concepts in civil aviation. Fault prediction-based decision support is expected to optimise the effectiveness and efficiency of maintenance. Thus, it also enables further reduction of an airline's operating costs over the aircraft life cycle. Prediction aims to transform unscheduled maintenance events, often causing operational irregularities, into projectable preventive activities. Thereby, aircraft availability as well as maintenance processes are expected to be optimised. Because of low technology readiness levels in the industry as well as rather global scale scientific publications on predictive maintenance, the presented work aims to analyse its implementation's potentials in a more detailed manner. Thus, a decision-supporting tool for the cost-benefit assessment of predictive maintenance opposed to the initial state is to be developed.
The starting point is literature research with respect to today's standards and characteristics in civil aviation aircraft maintenance. This includes the applied maintenance strategies and cost structures in particular. Thereafter, the state-of-the-art concerning fault prediction and relevant performance metrics is presented.
Subsequently, the proposed evaluation concept is introduced. Two functions are to be covered: Firstly, based on information on today's maintenance, cost reduction potentials as well as minimum requirements with respect to prediction can be derived. Secondly, prediction concepts can be assessed concerning their specific cost-benefit characteristics. Whereas the evaluation focus lies on the effected costs, corresponding time- and ratio-based target values are analysed as well. A unique feature of the method is the use of mostly deterministic data enabling the derivation of more accurate and valid results than probabilistic approaches.
Thereafter, the proposed model's software implementation is described. Based on theoretically defined business process and data models, a simulation model is built. The proposed model accounts for prediction-induced modifications within the aircraft maintenance as well as the interdependencies of the aircraft operations. By means of Monte-Carlo simulation, input data uncertainties are accounted for and processed for a statistical results assessment.
In a case study, results of the method's application are presented. Firstly, the calibration of estimated process efforts by means of available real-world maintenance information is conducted. This enables the model validation as well. Thereafter, the analysis results concerning an exemplary aircraft component that is maintained correctively are presented. This real-world data is then counterposed to target values derived from the simulation of prediction-based maintenance approaches.
It can be concluded that a predictive maintenance strategy's benefit depends on the amount or the ratio of the interdependent prediction errors, the prediction forecast as well as the costs of implementation. Among the prediction errors, it is necessary to distinguish between errors negatively affecting aircraft operations and errors having negative impact on aircraft maintenance activities. A longer forecast increases the ability to plan in advance, while also leading to higher prediction error rates. It is shown how the overall cost-benefit is affected by investment and operating costs of a predictive strategy's implementation. Only when the derived break-even thresholds are underrun, will the cost-benefit turn out positive as compared to the initial-state maintenance. The overall optimum incorporates a prediction model with the least costly parameter setting.
Typ des Eintrags: | Dissertation | ||||
---|---|---|---|---|---|
Erschienen: | 2017 | ||||
Autor(en): | Kählert, Alexander | ||||
Art des Eintrags: | Erstveröffentlichung | ||||
Titel: | Specification and Evaluation of Prediction Concepts in Aircraft Maintenance | ||||
Sprache: | Englisch | ||||
Referenten: | Klingauf, Prof. Dr. Uwe ; Metternich, Prof. Dr. Joachim | ||||
Publikationsjahr: | 2017 | ||||
Ort: | Darmstadt | ||||
Datum der mündlichen Prüfung: | 18 Januar 2017 | ||||
URL / URN: | http://tuprints.ulb.tu-darmstadt.de/6065 | ||||
Kurzbeschreibung (Abstract): | The goal of this thesis is to identify and quantify the potentials of predictive maintenance concepts in civil aviation. Fault prediction-based decision support is expected to optimise the effectiveness and efficiency of maintenance. Thus, it also enables further reduction of an airline's operating costs over the aircraft life cycle. Prediction aims to transform unscheduled maintenance events, often causing operational irregularities, into projectable preventive activities. Thereby, aircraft availability as well as maintenance processes are expected to be optimised. Because of low technology readiness levels in the industry as well as rather global scale scientific publications on predictive maintenance, the presented work aims to analyse its implementation's potentials in a more detailed manner. Thus, a decision-supporting tool for the cost-benefit assessment of predictive maintenance opposed to the initial state is to be developed. The starting point is literature research with respect to today's standards and characteristics in civil aviation aircraft maintenance. This includes the applied maintenance strategies and cost structures in particular. Thereafter, the state-of-the-art concerning fault prediction and relevant performance metrics is presented. Subsequently, the proposed evaluation concept is introduced. Two functions are to be covered: Firstly, based on information on today's maintenance, cost reduction potentials as well as minimum requirements with respect to prediction can be derived. Secondly, prediction concepts can be assessed concerning their specific cost-benefit characteristics. Whereas the evaluation focus lies on the effected costs, corresponding time- and ratio-based target values are analysed as well. A unique feature of the method is the use of mostly deterministic data enabling the derivation of more accurate and valid results than probabilistic approaches. Thereafter, the proposed model's software implementation is described. Based on theoretically defined business process and data models, a simulation model is built. The proposed model accounts for prediction-induced modifications within the aircraft maintenance as well as the interdependencies of the aircraft operations. By means of Monte-Carlo simulation, input data uncertainties are accounted for and processed for a statistical results assessment. In a case study, results of the method's application are presented. Firstly, the calibration of estimated process efforts by means of available real-world maintenance information is conducted. This enables the model validation as well. Thereafter, the analysis results concerning an exemplary aircraft component that is maintained correctively are presented. This real-world data is then counterposed to target values derived from the simulation of prediction-based maintenance approaches. It can be concluded that a predictive maintenance strategy's benefit depends on the amount or the ratio of the interdependent prediction errors, the prediction forecast as well as the costs of implementation. Among the prediction errors, it is necessary to distinguish between errors negatively affecting aircraft operations and errors having negative impact on aircraft maintenance activities. A longer forecast increases the ability to plan in advance, while also leading to higher prediction error rates. It is shown how the overall cost-benefit is affected by investment and operating costs of a predictive strategy's implementation. Only when the derived break-even thresholds are underrun, will the cost-benefit turn out positive as compared to the initial-state maintenance. The overall optimum incorporates a prediction model with the least costly parameter setting. |
||||
Alternatives oder übersetztes Abstract: |
|
||||
URN: | urn:nbn:de:tuda-tuprints-60655 | ||||
Sachgruppe der Dewey Dezimalklassifikatin (DDC): | 600 Technik, Medizin, angewandte Wissenschaften > 620 Ingenieurwissenschaften und Maschinenbau 600 Technik, Medizin, angewandte Wissenschaften > 650 Management |
||||
Fachbereich(e)/-gebiet(e): | 16 Fachbereich Maschinenbau 16 Fachbereich Maschinenbau > Fachgebiet für Flugsysteme und Regelungstechnik (FSR) |
||||
Hinterlegungsdatum: | 09 Apr 2017 19:55 | ||||
Letzte Änderung: | 09 Apr 2017 19:55 | ||||
PPN: | |||||
Referenten: | Klingauf, Prof. Dr. Uwe ; Metternich, Prof. Dr. Joachim | ||||
Datum der mündlichen Prüfung / Verteidigung / mdl. Prüfung: | 18 Januar 2017 | ||||
Export: | |||||
Suche nach Titel in: | TUfind oder in Google |
Frage zum Eintrag |
Optionen (nur für Redakteure)
Redaktionelle Details anzeigen |