As the volume of telemetry data generated by satellites and other complex systems continues to grow, there is a pressing need for more efficient and accurate anomaly detection methods. Current techniques often rely on human analysis and preset criteria, presenting several challenges including the necessity for expert interpretation and continual updates to match the dynamic mission environment. This paper critically examines the use of deep anomaly detection (DAD) methods in addressing these challenges, evaluating their efficacy on real-world spacecraft telemetry data. It exposes limitations in current DAD research, highlighting the tendency for performance results to be overestimated and suggesting that simpler methods can sometimes outperform more complex DAD algorithms. By comparing established metrics for anomaly detection with newly proposed ones, this paper aims to improve the evaluation of DAD algorithms. It underscores the importance of using less accuracy-inflating metrics and offers a comprehensive comparison of DAD methods on popular benchmark datasets and real-life satellite telemetry data. Among the DAD methods examined, the LSTM algorithm demonstrates considerable promise. However, the paper also reveals the potential limitations of this approach, particularly in complex systems that lack a single, clear predictive failure channel. The paper concludes with a series of recommendations for future research, including the adoption of best practices, the need for high-quality, pre-split datasets, and the investigation of other prediction error methods. Through these insights, this paper contributes to the improved understanding and application of DAD methods, ultimately enhancing the reliability and effectiveness of anomaly detection in real-world scenarios.

Over the past years, advanced prognostic models and approaches have been developed. Most existing approaches are tailored to one specific system and cannot adaptively be used on different systems. This can lead to years of research and expertise being put into implementing prognostic models without the capacity to predict system failures, either because of a lack of data or data quality or because failure behavior cannot be captured by data-driven models. In addition, prognostic models are often evaluated using metrics only related to the correctness of predictions, preventing meaningful evaluation of operational performance. This paper makes use of a framework that can automatically choose prognostic settings based on specific system data. It simultaneously optimizes the choice of methodologies using metrics that capture multiple aspects of prediction quality. We apply this framework to both a simulated data set and a real aircraft data set to characterize the impact of metrics on the choice of prognostic methodologies. The results show that the choice of optimization metric greatly impacts the output of the generic prognostic framework and the overall performance. In addition, a definition for data suitability is provided and assessed on the aircraft system data sets.

Prognostics and Health Management (PHM) models aim to estimate remaining useful life (RUL) of complex systems, enabling lower maintenance costs and increased availability. A substantial body of work considers the development and testing of new models using the NASA C-MAPSS dataset as a benchmark. In recent work, the use of ensemble methods has been prevalent. This paper proposes two adaptations to one of the best-performing ensemble methods, namely the Convolutional Neural Network - Long Short-Term Memory (CNN-LSTM) network developed by Li et al. (IEEE Access, 2019, 7, pp 75464-75475)). The first adaptation (adaptable time window, or ATW) increases accuracy of RUL estimates, with performance surpassing that of the state of the art, whereas the second (sub-network learning) does not improve performance. The results give greater insight into further development of innovative methods for prognostics, with future work focusing on translating the ATW approach to real-life industrial datasets and leveraging findings towards practical uptake for industrial applications.

Aircraft fleet managers lack tools to aid decision-making for fleets nearing retirement, which leads to rushed and ill-informed decisions. Accordingly, aging aircraft fleets are underutilized and fleets can be retired before their useful lifetime has been expended. A decision support framework is proposed to solve the aging military aircraft retirement problem. It integrates four steps for fleet managers to simplify the decision-making process: (i) Understanding the structural toll caused by utilization, (ii) Recognizing the indicators that predispose a fleet for retirement, (iii) Determining an optimal fleet size and choosing which aircraft to retire and (iv) Optimizing end-of-life usage prior to retirement. An example using a sample military fleet is used to illustrate the effectiveness of the decision support framework, integrating both computational results and manager judgement. Fleet managers were used to validate the concepts in the framework and their opinions are presented herein. It is shown that fleet managers can utilize a decision support framework to positively impact their decision-making for full-spectrum aging aircraft retirement decisions.

During flights aircraft continuously collect data regarding operations, health status and system condition. Data-driven approaches typically applied to system specific sensor data provide a way to predict failures of aircraft systems. However, it is believed that some systems deteriorate faster when subjected to particular environmental conditions, such as humidity or dust. In this study, we consider an aircraft system which is suspected to experience degradation due to humidity during ground operations. We apply a Random Forest approach to sensor data only and a combination of sensor data and environmental data from airports to estimate the system's remaining useful life. To our knowledge this is the first paper addressing the problem of integrating environmental data in prognostics for aircraft systems using raw sensor data. The method is validated on a data set provided by an airline that includes the per-second sensor data of 11 different sensors for roughly 12,300 flights, as well as 15 removals. Meteorological data for airports worldwide is obtained from the Meteorological Aerodrome Reports database. The results show that incorporating environmental data in prognostics has a potential towards more accurate prediction models.

Aircraft dispatch involves determining the optimal dispatch option when an aircraft experiences an unexpected failure. Currently, maintenance technicians at the apron have limited access to support information and finding the right information in extensive maintenance manuals is a time-consuming task, often leading to technically induced delays. This paper introduces a novel web-based prototype decision support system to aid technicians during aircraft dispatch decision-making and subsequent maintenance execution. A system architecture for real-time dispatch decision support is established and implemented. The developed system is evaluated through a case study in an operational environment by licensed maintenance technicians. The system fully automates information retrieval from multiple data sources, performs alternative identification and evaluation for a given fault message, and provides the technician with on-site access to relevant information, including the related maintenance tasks. The case study indicates a potential time saving of up to 98% per dispatch decision. Moreover, it enables digitalization of the—currently mostly paper-based—dispatch decision process, thereby reducing logistics and paper waste. The prototype is the first to provide operational decision support in the aircraft maintenance domain and addresses the lack of correlation between theory and practice often found in decision support systems research by providing a representative case study. The developed custom parser for SGML-based documents enables efficient identification and extraction of relevant information, vastly contributing to the overall reduction of the decision time.

This paper proposes an analytical model that uses historical damage dimension data to deduce physical impactor characteristics (size and energy) that has caused a certain resulting damage. Maintenance tasks occur in operations due to impact, however the source of the damage caused in the event remains in most cases unknown. Consequently, by inferring what has caused a certain type of damage from the distribution of the damage type and severity relative to impactor types, maintainers can be better prepared in terms of what to expect from a given impactor source. The developed model introduces a novel transition deformation region between the local deformation and the global plate deflection, allowing for fast and accurate predictions of the impact event. Using the known aluminium structural properties and damage dimensions, the damage data is converted into impactor data. The model is applied in a case study using 120 fuselage dent damages dimensions (length, width, and depth) from a Boeing 777 fleet. The results show that the model deduces impactor characteristics for 94% of the considered damages, ranging up to 240 J and 110 mm for impactor energy and radius respectively.

Prognostic and health management (PHM) describes a set of capabilities that enable to detect anomalies, diagnose faults and predict remaining useful lifetime (RUL), leading to the effective and efficient maintenance and operation of assets such as aircraft. Prior research has considered the methodological factors of PHM system design, but typically, only one or a few aspects are addressed. For example, several studies address system engineering (SE) principles for application towards PHM design methodology, and a concept of requirements from a theoretical standpoint, while other papers present requirement specification and flow-down approaches for PHM systems. However, the state of the art lacks a systematic methodology that formulates all aspects of designing and comprehensively engineering a PHM system. Meanwhile, the process and specific implementation of capturing stakeholders’ expectations and requirements are usually lacking details. To overcome these drawbacks, this paper proposes a stakeholder-oriented design methodology for developing a PHM system from a systems engineering perspective, contributing to a consistent and reusable representation of the design. Further, it emphasizes the process and deployment of stakeholder expectations definition in detail, involving the steps of identifying stakeholders, capture their expectations/requirements, and stakeholder and requirement analysis. Two case studies illustrate the applicability of the proposed methodology. The proposed stakeholder-oriented design methodology enables the integration of the bespoke main tasks to design a PHM system, in which sufficient stakeholder involvement and consideration of their interests can lead to more precise and better design information. Moreover, the methodology comprehensively covers the aspects of traceability, consistency, and reusability to capture and define stakeholders and their expectations for a successful design.

Aircraft maintenance has been further developed with predictive maintenance instead of solely condition-based maintenance. Prognostics and health management (PHM) with advanced technologies can utilize real-time and historical health state information to provide actionable information, enabling predictive maintenance decision-making. In this case, the methodology of how to design the PHM systems is an issue to be faced. The state of the art has provided several conceptual design methodologies and associated methods to support the conceptual requirements development of PHM systems. However, there is no rigorous process available for requirements definition. Existing options for requirements derivation are lacking details, which restricting PHM system design and development. This constitutes a major drawback and hurdle towards the successful design of PHM systems in practice. This paper consequently proposes a methodology for the systematic derivation of system requirements towards PHM system development. Besides, this methodology defines detailed processes for requirements definition, and positions mean through which various categories of requirements can be derived through appropriate analyses in detail. Sequences of interoperability requirements categories and associated flow-down perspectives are identified. To evaluate the applicability, this paper undertakes the case study of requirements definition for a generic PHM system, which provides a comprehensive application of the methodology. Designers can perform requirements definition under this methodology as guidance towards the design of a successful PHM system, providing solutions for predicting remaining useful life (RUL) to support aircraft predictive maintenance.

When an aircraft experiences an unexpected issue during flight operations, a technician determines whether the aircraft can safely perform the next flight. This operational decision process - known as dispatch assessment - has to happen within limited available time between aircraft arrival and departure. Currently, technicians face two main problems during the assessment: lack of access to decision support information and a time-consuming process for finding relevant information in extensive maintenance manuals. These issues often lead to delays and additional costs and are indicative of three larger challenges in the decision support domain: 1) a paucity of decision support models and applications for operational processes in maintenance; 2) relatively few efforts in applying and evaluating artifacts in experimental and real-life operational settings; and 3) a lack of systematic development, application and evaluation of digitization and automation efforts of complex decision processes in maintenance. This paper applies a design science research approach to address these challenges and introduces two novel artifacts: a decision support framework for real-time decision making in aircraft dispatch, and a web-based prototype tool accessible through mobile solutions. The practical relevance of the framework and prototype is validated through two representative application and evaluation studies, one in an experimental setting and one in an operational environment. Results show significant time savings and strong qualitative indications towards a higher incentive to use documentation and reducing human risk factors that lead to maintenance error.

The system concept has existed for several decades now, but is still a viable concept to be used to denote a problem area and to adopt a holistic view. The essence of a system is that it consists of elements and relationships between these elements, and that it exerts a function in its environment, provided it is an open system. A system can be defined at different layers of abstraction consisting of subsystems, which themselves may consist of subsystems again. Themost complex level includes human beings. The system concept is adopted in Systems Engineering (SE) in which not only the engineering system under development is modeled, but also the development process itself in which many different disciplines need to be involved depending on the (lifecycle) requirements in focus. In this introductory chapter we draw the way we have paved to provide this book from the first idea on. The system concept, the origins, the goals and the expected audience of this book are roughly described. Finally, we give the first insight in the structure of this book and themutual interdependence of the chapters. This book contains many different contributions in the area of SE, categorized into 4 parts: an introduction to the concept, methods and tools, applications, and challenges.