Scheduled aircraft maintenance consists of a set of work tasks to be executed. This thesis focuses on improving the order in which these work tasks are executed. Because finding an optimum for these kind of problems is computationally intensive, the chosen approach is of heuristic nature, which is a practical method resulting in close-to-optimal solutions. In short, the model builds a schedule based on a set of rules and constraints, in which the work tasks with the highest priority are planned as early as possible in the maintenance check. Here, priority is the heuristic and based on minimizing the total length of the maintenance check. Because of the rule-based approach, the output of the model will only slightly change with similar sets of work tasks, leading to standardization of the maintenance. Lastly, as the model is semi-automatized, the model can both proactively as well as reactively generate schedules.

Delivering air power is the main goal of an Air Force. Military aircraft used to deliver this air power are required to be maintained in accordance with a so-called Aircraft Maintenance Program (AMP). Utilization of military aircraft is assumed to be constant when constructing an AMP. In reality, usage can drastically change over both short- and long-term periods as a result of political choices, as well as sudden deployments and changing mission types. These changes in usage will have its effect on system condition and by extension on required maintenance. In order to guarantee optimal efficiency and effectiveness of the AMP, these changes in usage should be taken into account while reviewing the AMP. Health and Usage Monitoring Systems (HUMS) can be used to support in analysing usage on a detailed level. A tool which is able to incorporate both historical and (expected) future usage will be of great help to a maintenance engineer responsible for reviewing the AMP on a yearly basis. In this research, a tool has been developed able to support in maintenance decisions on interval adjustments and incorporating system usage. Various methods and algorithms are incorporated of which some are novel contributions in their own right. The developed tool uses HUMS data, flight planning data, and design data to provide a possible adjustment on maintenance intervals. The tool takes both historical usage and expected future usage into account. The model is able to translate usage profiles (as defined by the operator) into usage on a very detailed level. This translation creates possibilities to determine usage effects on system condition. Translation to this detailed level of usage is achieved by processing HUMS data through rule-based flight regime recognition algorithms. Finally, the model includes a method to carry out a simplified business case to show possible gains resulting from the proposed interval adjustment. The model is tested by applying case study data originated from the Apache helicopter of the Royal Netherlands Air Force (RNLAF). The used data to train and test the model covers three years of operation: 2014-2016. Concerning accrued fatigue damage, case study results show an average severity factor of 0.37. This factor is based on the accrued fatigue damage resulting from usage conform the design spectra as defined by the Original Equipment Manufacturer (OEM). The maintenance intervals, currently used by the RNLAF and defined in the AMP, are based on these design spectra. Validation of the model is carried out by making use of a dedicated validation dataset (covering the year 2017). Both historical and expected (extreme severe) future usage are incorporated to set up business cases for three Fatigue Life Limited (FLL) Critical Safety Items (CSIs). These business cases revealed potential gains in component and maintenance cost when adapting the proposed interval adjustment. For these three components, the model proposed interval escalations of 42%, 44%, and 94% where both historical and expected future usage is taken into account.

Air forces demand aircraft availability for their operations. In order to increase this availability, there is a need for insight in the effect of environmental conditions on the usage profiles of components and systems. Existing frameworks and standards only provide for static environmental effects on component degradation. The research presented in this thesis aims to identify primary environmental drivers of deterioration for selected systems and components by modeling, implementing and subsequently verifying and validating these drivers on maintenance requirements. In this research, a multi-level prognostic framework is proposed. The framework is tested and applied to a case-study within the Royal Netherlands Air Force at the NH90 program. Using Principle Component Analysis, coefficients for the first two components show that Relative Humidity, Sea Level Pressure and Temperature-related features are environmental drivers to the corrosion degradation characteristics of the NH90. These results are verified on the maintenance requirements of the RNLAF, which show an increase in system availability and less maintenance costs such as spare parts demand and required personnel.

Unplanned maintenance is a costly factor in aircraft operations. Predictive Maintenance aims at reducing the surprise effect of unplanned maintenance and thereby its associated cost. A variety of statistical models are used to estimate the remaining life, as well as sensors to gauge component condition. The application to statistical models of sensory information coming from the pilot, in the form of pilot complaints, appears to be an overlooked option worth investigating. In other words: What is the effect of pilot complaints on the predictability of component removals? This question is answered by determining relevant words in the pilot complaints using a TF-IDF analysis and use the presence of these words as covariate in the well known Proportional Hazards Model. Left truncation and right censoring is applied to limit the time-invariant nature of these covariates. The results in the form of hazard ratios indicate a hazard increase of several orders of magnitude with respect to baseline hazard. These results are put into perspective when compared when compared to the known outcome of the pilot complaints, making their added predictability seem marginal. Another adverse indication is the violation of the proportionality assumption. The magnitude of the hazard ratios do suggest that additional measures in the from of a more in depth natural language processing and the application of time-varying covariates could bring the concept closer to practical application.

Aircraft maintenance is inherently unpredictable due to differences in failure rates between components. This results in high variance inflow in repair shops, which in case of insufficient capacity yields long turnaround times and thus the need for additional components in stock. The aim of this research project is to determine the effect of service level optimisation in the presence of highly variable demand from an aircraft maintenance supply chain perspective. The first approach focuses on determining the optimal capacity in the repair shop using a greedy algorithm, while the second aims at controlling demand by assisting in in- or outsourcing decision-making by means of a multi-criteria decision-making model. A case study at KLM Engineering & Maintenance is performed to test the applicability of the models in the aircraft maintenance industry.

Aircraft base maintenance consists of hundreds of tasks including preventive, corrective and modification tasks which have to be scheduled to a few check moments per year. The allocation of tasks to checks represents a complex scheduling problem due to the constraints involved. A mixed integer linear programming model was developed to optimize the allocation of the complete base maintenance workload and maintenance checks over a four year period. The model optimizes for resource capacity and interval efficiency. CPLEX has been used as solver. An iterative solution technique loops around the model which yields an approximate optimal solution. The model was developed for a large MRO service provider. Its outcome provides the MRO with an integral insight into the complete workload distribution over a tactical planning horizon for all checks. This insight can help to optimize the allocation of maintenance tasks and resource capacity which will contribute to a higher fleet availability.

Airlines experience schedule disruptions on a daily basis. Poor weather conditions, unscheduled aircraft maintenance and congested air spaces are just a few of the causes that prevent airlines from operating their flight schedules as planned. In the third quarter of 2017 over 20% of all scheduled flights in Europe suffered from delays. Operation Research based decision support systems (DSS) help airlines with their disruption management processes and provide suggestions for recovery options. For large hub-and-spoke carriers, with an extensive network and a large number of aircraft and flights, the computation time required to find the optimal recovery solution after a disruption increases rapidly. As airlines require fast recovery solutions when disruptions occur, there is an ongoing trade-off between computation time and system sophistication. In the majority of disruption cases, no or a limited number of undisrupted aircraft are required to find the optimal recovery solution. By selecting a limited number of aircraft, flights and airports used to find a recovery solution, the computation time is reduced exponentially. The challenge is determining which aircraft and flights should be selected. This research aims to develop a decision support system for the schedule and aircraft recovery process that is able to present a feasible solution to a disruption in less than 120 seconds. An aircraft recovery model will be developed based on the integer linear programming model that was created by Vink et al. (2019) and Vos et al. (2015). Crew and passenger recovery are not considered. To recover disruptions, the optimization model can delay and cancel flights as well as perform tails swaps, where the flights from two aircraft are switched. The novelty of the work is that machine learning is used to predict which undisrupted aircraft will help recover a disruption. Based on those predictions a sub-network selection algorithm will select the subset of aircraft to be included in the optimization instead of the entire aircraft fleet. The performance of the system is tested on a case study for the domestic hub-and-spoke network of Delta Airlines. The dataset for the study consists of 2200 daily flights, 147 airports and 827 aircraft in 8 aircraft families. The results of the system are compared with the optimal solution, where no aircraft selections were made. The case study shows that the system is able to make an aircraft selection where 50% of the fleet is discarded, while still finding the optimal solution in 98.9% of the 556 disruptions tested. Furthermore, the system reduces the computation time by 45%, resulting in an average time of 48 seconds. For the disruptions, the computation time varied between 9 and 180 seconds.

The newest generation of aircraft has seen a strong increase in sensor data generated on-board. The available data has the potential to indicate the health state of individual components based on which their maintenance requirements can be determined, a maintenance strategy called Condition Based Maintenance. Predictive Maintenance is a specific condition based maintenance strategy that aims to determine these requirements in advance by predicting failures from the sensor data. It has the potential to reduce costly unanticipated maintenance or unnecessarily conservative maintenance. In the short term it could add significant value for components that are currently subject to a reactive maintenance policy. In the long term it could potentially disrupt the traditional maintenance practice of periodic inspection. One of the main challenges in applying Predictive Maintenance in the aviation industry is translating the large amounts of sensor data into a reliable failure prediction, a process called prognostics. In this study, state-of-the-art machine learning, and specifically deep learning models, have been investigated for their potential for prognostics. A case study has been performed at KLM Royal Dutch Airlines on the Boeing 747 Bleed Air Valves, traditionally some of the most challenging components from a maintenance perspective. It has been shown that fully self-learning algorithms can be used for prognostics, enabling the implementation of one of the first real-life predictive maintenance implementations.

In aircraft maintenance, performance of routine maintenance tasks is prone to generate additional non-routine tasks. These non- routine tasks relate to solving an unexpected issue and have specific requirements in terms of materiel, tools and manpower. In existing research regarding optimization of aircraft maintenance environments and schedules, these requirements are assumed to be in place or not considered at all. In real-life, the required resources may not be in place and retrieving them is subject to human interactions and decision-making, potentially resulting in delays in the maintenance schedule. This research proposes a novel methodology to address this problem by developing a multi-agent automated negotiation model capable of simulating negotiations related to the procurement of required aircraft materials in order to perform non-routine tasks. The proposed model explores its applicability by incorporating several negotiation strategies, adaptive parameters and negotiation circumstances, e.g. urgency. The main research question is: In the context of non-routine aircraft maintenance materiel availability, to what extent can multi- agent automated negotiation be applied to explore effective agent strategies for various negotiation circumstances regarding the procurement of aircraft components and expendables? The proposed methodology has shown to be very suitable for modelling negotiation circumstances as evaluation of the model’s performance has resulted in understanding and recommendations to policy makers concerning what strategy to apply in which circumstance. Variations of circumstances have lead to expected findings, e.g. increasing urgency results in a sooner delivery of materials, whereas in other situations the model’s KPIs are unaffected by parameter variations, unexpectedly.

Wind energy is one of the fastest growing types of renewable energy. Today, offshore wind represents 14% of the EU wind energy market, and has a huge potential for further growth [1]. Operations and Maintenance (O&M) accounts for 18-23% of the total lifetime cost in offshore wind farms, compared to 12% onshore [2]. The offshore O&M tasks are more costly, being influenced by distance offshore, harsh offshore conditions, wind farm size, wind turbine reliability and maintenance strategy [2]. Exchanges of main components, also known as major interventions, causes significant costs and has been indicated as the largest uncertaintywhen predicting O&M costs [3]. Major interventions are characterized by the need for a jack-up vessel to perform the exchange. No studies have been found on offshore major interventions. Echavarria has previously performed an analysis on the exchange rates of main components based on the onshore WMEP population, which had an average rated power of 233 kW [4]. Blades and generator were identified as most critical components, but it is not known how this reflects current exchange rates for larger turbines. The goal of this research is therefore to determine future failure rates and predict the cost of offshore major interventions.

There is a growth in the use of composites for the new generation of wide-body aircraft such as the Boeing 787 and Airbus A350. This shift from using aluminium as the primary material is motivated by the benefits of using composites in design, manufacturing and operations. Composites offer the aircraft manufacturer the ability to create more complex shapes and optimise the design such that it is light-weight. This, in tandem with other design improvements, leads to lower fuel burn. Consequently, airlines see the advantage of these new aircraft to reduce their operational cost. Therefore, as airlines continue to renew their ageing fleets of aluminium aircraft, there is going to be an increased need for composite maintenance. However, fulfilling the increased demand for composite repairs is impeded by limited availability of historical damage data, due to the young operational age of these aircraft. Composites are particularly sensitive to impact damage, and understanding the likelihood and the consequence of this type of damage is valuable for maintenance processes such as repair decision-making. The purpose of this dissertation is to predict the risk of impact damage for future composite aircraft and use it to substantiate maintenance decision-making in an operational setting...@en

Airline Maintenance and Engineering (M&E) organizations are faced with a number of repairs time to time for their fleet of aircraft due to accidental damages. As these damages are unpredictable in nature, the approach to repairing these damages is reactive. These type of repairs fall under the category of corrective or unscheduled maintenance policy compared to the planned preventive or scheduled maintenance. As the occurrence of these unscheduled repairs result in consumption of more maintenance resources in an untimely manner, they add to the existing costs for the organisation. Hence, it is of interest to the M&E to predict the demand for these resources for a future period so that the organisation is better prepared to handle future maintenance activities. One of the resources impacted due to unscheduled repairs is capacity, i.e maintenance hangar facility. If the capacity of a hangar facility can be divided into certain number of slots, then prediction of the demand for these slots would be beneficial to the maintenance planner. In order to identify the demand for these slots, it is first important to forecast the trend of unscheduled repairs. To achieve this goal, i.e. prediction of future repair and determination of slot capacity, a novel application for the integrated use of a reliability and inventory control model has been identified in this thesis. Here, the concepts of inventory control has been specifically applied to a maintenance application to determine the maintenance capacity by taking into account the stochastic demand of unscheduled repairs. The model used to predict the demand of unscheduled repair is a Non-homogeneous Poisson Process (NHPP) reliability model with a Power law intensity function and the inventory control model that was found to be applicable is the singlesystem single location Base-stock policy model. The reliability model considers the superpositioning principle through which the failure behaviour for the entire fleet of aircraft could be predicted. Certain performance measures were identified from the inventory control model, which helped in determining capacity based on optimum costs as well as service level requirements. As a proof of concept, a study is done on identifying the long-run capacity requirements for a fleet of Boeing 777 aircraft of a major European airline. Two specific structural components were identified on which the study was carried out, namely, the leading slats and the outboard flaps. The results showed the successful implementation of the model by identifying 30 slots necessary in the next 1500 flight cycles at an optimum cost for the case of leading edge slats.