Archive ouverte de l'ENAC
Not a member yet
3458 research outputs found
Sort by
Principle of TEM alignment using convolutional neural networks: Case study on condenser aperture alignment
No full textInternational audienceThe possibility of automatically aligning the transmission electron microscope (TEM) is explored using an approach based on artificial intelligence (AI). After presenting the general concept, we test the method on the first step of the alignment process which involves centering the condenser aperture. We propose using a convolutional neural network (CNN) that learns to predict the x and y-shifts needed to realign the aperture in one step. The learning data sets were acquired automatically on the microscope by using a simplified digital twin. Different models were tested and analysed to choose the optimal design. We have developed a human-level estimator and intend to use it safely on all apertures. A similar process could be used for most steps of the alignment process with minimal changes, allowing microscopists to reduce the time and training required to perform this task. The method is also compatible with continuous correction of alignment drift during lengthy experiments or to ensure uniformity of illumination conditions during data acquisition
A Review of Situational Awareness in Air Traffic Control
No full textInternational audienceAs demand for air transport continues to grow, the complexity of the operating environment of air traffic has intensified. Air Traffic Controllers (ATCos) are faced with the critical task of making timely decisions in response to rapidly changing air traffic. In this context, maintaining Situation Awareness (SA) becomes critical, directly influencing ATCos' decision making and preventing potential traffic accidents or incidents. This paper presents a systematic review of the theoretical development of SA measurement techniques for ATCos. Firstly, the measurement techniques developed to assess individual and team SA are discussed and summarized. Additionally, some recently developed novel measurement methods are introduced. Four specific techniques applied to evaluate ATCos' SA are highlighted. Second, the article analyzes the comprehensive utilization of neurophysiological measurement techniques, with a particular focus on eye movement and EEG methodologies. These techniques exhibit promising potential in measuring ATCos' SA and are explored in-depth. Furthermore, four distinct types of sensitivity factors that primarily affect ATCos' SA are summarized. This paper provides some insight into the current state of SA measurement techniques for ATCos. It concludes with recommendations for future research, specifically addressing evaluation of ATCos SA in high-automation environments.</div
Two-stage Stochastic Optimization for the Extended Aircraft Arrival Management Problem Under Uncertainty
No full textInternational audienceThis chapter reviews recent developments to manage aircraft arrivals in the context of extended arrival manager systems, for which uncertainty is significant when predicting expected times to start the approach phase and landing times. An original high-level multi-stage stochastic optimization formulation, considering several air network points of interest, is first introduced, taking account of practical operational constraints. The remaining of the chapter focuses on the two-stage special case, which corresponds to recent studies on the aircraft arrival management problem. A landing order is decided at a specific air network point known as the initial approach fix, or IAF (first stage), and a recourse cost is proposed so as to ensure that aircraft separation constraints are satisfied at the landing runway (second stage). Multiple possible IAF points are considered as well as the possibility to delay the departure of on-ground aircraft. Finally, this study proposes new analyses (validation score and impact of inclusion of chance constraints in the first stage) of numerical experiments performed on realistic instances based on Paris-Charles de Gaulle arrival data. We discuss numerical results and exhibit that the stochastic solutions are more robust than their deterministic counterparts.</div
Articulating Safety and Climate Change
No full textInternational audienceAbstract With the increasing number of climate-related events and the growing public awareness of climate change and its potential consequences, new challenges emerge including for high-risk industries and the way to manage their safety. Ignoring the interrelations between the two societal stakes that are climate change and safety might lead to critical situations. This chapter identifies the questions raised by addressing the interplay between safety and climate change at both conceptual and practical levels. It also suggests perspectives for safety science and scientists to have a more significant contribution in this unprecedented context
A Novel Aircraft Trajectory Generation Method Embedded with Data Mining
No full textInternational audienceData mining has achieved great success in air traffic management as a technology for learning knowledge from historical data that benefits people. However, data mining can rarely be embedded into the trajectory optimization process since regular optimization algorithms cannot utilize the functional and implicit knowledge extracted from historical data in a general paradigm. To tackle this issue, this research proposes a novel data mining-based trajectory generation method that is compatible with existing optimization algorithms. Firstly, the proposed method generates trajectories by combining various maneuvers learned from operation data instead of reconstructing trajectories with generative models. In such a manner, data mining-based trajectory optimization can be achieved by solving a combinatorial optimization problem. Secondly, the proposed method introduces a majorization–minimization-based adversarial training paradigm to train the generation model with more general loss functions, including non-differentiable flight performance constraints. A case study on Guangzhou Baiyun International Airport was conducted to validate the proposed method. The results illustrate that the trajectory generation model can generate trajectories with high fidelity, diversity, and flyability
Arrival and Departure Sequencing, Considering Runway Assignment Preferences and Crossings
No full textInternational audienceAircraft sequencing has the potential to decrease flight delays and improve operationalefficiency at airports. This paper presents the aircraft sequencing problem (ASP) on multiple runwayswith complex interactions by allocating flights on runways and optimizing landing times, take-offtimes, and crossing times simultaneously in a uniform framework. The problem was formulated as amixed-integer program considering realistic operational constraints, including runway assignmentpreferences based on the entry/exit fixes of the terminal maneuvering area (TMA), minimum runwayseparation, time window, and arrival crossing rules. Variable-fixing strategies were applied, tostrengthen the formulation. A first-come-first-served (FCFS) heuristic was proposed for comparison.Various instances from the literature and realistic data sets were tested. Our computationalstudy showed that the solution approach optimizes runway schedules, to achieve significantly fewerflight delays, taking runway assignment preferences and arrival crossings into accoun
Decision-making under environmental complexity: the need for moving from avoided impacts of ICT solutions to systems thinking approaches
No full textInternational audienceThe indirect impacts of Information and Communication Technology (ICT) on the environment (whether positive or negative) have been extensively discussed in academic and industrial literature, particularly within the ICT4S community. However, a lack of consensus exists in academia on how to assess them, especially in the context of decision-making processes. This paper examines whether ‘net impacts accounting’ methods are suitable for decision-making and suggests alternative approaches. We begin by clarifying different scenarios within the context of environmental decision- making. Then, we assess their relevance across those different decision scenarios. We emphasize their inadequate response to uncertainties, their focus on solutions rather than problems, and their inability to inspire a range of decisions compatible with an environmental transition. Drawing insights from systems thinking, we finally suggest methods and tools that could be combined to better address the complexity surrounding environmental decision-making. Throughout the paper, we develop the case study of Vinted – a second-hand clothing resale platform – to illustrate our arguments. The contribution advocates for a more systemic approach that embraces complexity by employing mixed methods, encompassing both qualitative and quantitative perspectives
Towards aircraft generic Quick Access Recorder fuel flow regression models for ADS-B data
No full textInternational audienceThis paper presents an investigation into the de-velopment of generic Quick Access Recorder (QAR) fuel flowregression models applied to Automatic Dependent Surveillance-Broadcast (ADS-B) data, with the aim of improving the accuracyof fuel flow estimates for various aircraft operations. Given thecritical need for accurate fuel consumption estimates to mitigatethe environmental impact of aviation, this study explores a novelapproach that integrates derivative features and aircraft-specificparameters into a unified model. This approach not only aimsto generalise across different aircraft types, thereby providingscalability and flexibility for end users, but also demonstratesadaptability to missing parameters through data augmentationtechniques. Using a dataset of QAR data from various aircraft type,this paper evaluates the performance of the model across differentflight phases and aircraft types and compares it with other commonfuel flow models
Flow based spatio-temporal graph network model for predicting airport network delays
No full textInternational audienceWith the continued advancement of the air transportindustry, ensuring the safety and efficiency of air traffic operationshas become a major concern. Various efforts are being made toachieve this objective, among which the prediction of delays inair traffic is of significance. Current deep learning methodologiesfor predicting network wide air traffic delays typically rely onhistorical delay data as the primary input, often neglecting theinclusion of air traffic flow data. In this paper, the FSTGMANmodel is developed to explore the efficacy of incorporating trafficflow data as an additional input for predictive modeling, contrastingits performance with a model that does not utilize flow data.The findings reveal that the incorporation of flow data marginallyenhances the overall accuracy of the predictions. Furthermore, theperformance of our model is compared with that of baseline modelssuch as MLP, LSTM, Transformer, and Seq2Seq, demonstratingnotable advantages
Crossing Waypoint Optimization in Free Route Airspace
No full textInternational audienceWith the imminent growth in air travel, ensuring theefficiency of airspace has become a necessity. This work addressesthis specific concern, namely, the optimization of a Free RouteAirspace (FRA). The FRA is a relatively recent concept appliedinto the european airspace. Even though this approach has theadvantage of having the direct routing between origin and destination, there are still aspects that can be improved, regardingconflict handling and air traffic controllers’ workload. To face thisproblem, we study the incorporation of Crossing Waypoints (CWs)into the FRA. These CWs will cluster the locations of conflicts,facilitating the controller’s task. However, identifying their bestlocations and connections poses a challenge. Along this research,the problem of finding the optimal topology will be modeled inorder to minimize potential conflicts between flights, reduce theworkload on controllers, and decrease airlines’ costs. This willlead to a multi-objective optimization problem, involving a balancebetween safety and fuel consumption. To find a solution to thisissue, meta-heuristics algorithms were employed and compared(Particle Swarm Optimization, Simulated Annealing and TabuSearch). Initially the model will be validated in simple cases inorder to compare the use of direct routing and the CW approach.Then the procedure is evaluated in a case-study of sector LFEEKFin Reims. Finally, we present a comparison between the algorithmsto evaluate their individual performances