Archive ouverte de l'ENAC
Not a member yet
3458 research outputs found
Sort by
Flight allocation in flight-centric air traffic control: Hierarchical clustering and Simulated Annealing approach
No full textInternational audienceThis study explores flight allocation within a flight-centric air traffic management framework using hierarchical clustering and Simulated Annealing. The proposed approach involves merging interacting flights before balancing the controllers' workload with Simulated Annealing. An analysis of the impact of the grouping threshold is conducted to identify the value that best balances minimizing interactions between flights assigned to different controllers and the controllers' workload. Finally, a comparison with a MILP model demonstrates that the new approach is more computationally efficient and more realistic from an operational perspective
Non-homotopic alternative short paths in the plane withpolygonal obstacles
No full textInternational audienceIn the context of geometric collision avoidance for drones and eVTOLs (electric VerticalTake-Off and Landing systems) in urban environments, we study the possibility of enhancinganticollision algorithms with strategic path planning [1]. Our intended approach involves calculating alternative paths through the obstacles (typically buildings) for each flying agent,assuming that the traffic is organized in horizontal layers. The focus of this paper is on findingdiverse short paths in the plane with obstacles.Classical techniques such as shortest-path or K-shortest-path algorithms often produce similar paths differing only by a few waypoints. To introduce diversity, some approaches addrandom obstacles [2] or use Yen’s algorithm and similarity metrics to speed up path finding[3]. In this paper, we use a topological definition of diversity based on homotopy equivalenceclasses. Two paths are homotopic if one can be continuously deformed into the other withoutcrossing obstacles. Previous works have explored this idea. The article [4] proposes to build agraph from the Voronoi diagram and then search for homotopically different paths with random walk exploration. [5] presents a mathematical method based on complex analysis to searchfor exact homotopy classes. Finally, in [6] the authors first construct the graph of the Voronoidiagram and then search for shortest paths inside it.Our proposed method first searches for homotopically different "funnels" in a trapezoidaldecomposition of the plane with obstacles. Here, a "funnel" is a sequence of connected trapezoidsconnecting the departure point to the arrival. An A∗ algorithm is then applied on a subgraphof the visibility graph, in order to find the shortest path in each selected funnel
Uncertainty Quantification of First Fix in a Time-Differenced Carrier Phase Observation Model
No full textInternational audienceThis paper presents an uncertainty quantification analysis of the first fix in a time-differenced carrier phase (TDCP) observation model. TDCP is a widely used method in GNSS-based odometry for precise positioning and displacement estimation. A key point in the TDCP modeling is the assumption that the GNSS receiver's initial position is perfectly known, which is never exactly the case in real-world applications. This study assesses the impact of initial position errors on estimated displacement by formulating a correct TDCP model and a misspecified one, where the first position is not correct. Theoretical derivations provide a generic framework of estimation under the misspecified model and its associated mean squared error (MSE), as well as estimation performance bounds through the misspecified Cramer Rao bound (MCRB) for the considered case. These theoretical considerations are then used to build an estimator of the receiver's displacement, with comparisons to the MCRB for performance evaluation. Extensive simulations using realistic GNSS geometry assess the influence of a first-fix error under various conditions, including different time intervals, first-fix error norms, and first-fix error direction. As an example, it is shown that for the considered geometry, if a TDCP of t_2 - t_1 = 1 s is built with an initial first fix error norm ∥∆r 1 ∥ = 10 m, then it introduces an estimation of the displacement, with an error of norm equal to 1.3 mm, at most. The results indicate that the displacement estimation error is linearly related to the initial position error and the time interval between observations, highlighting the importance of accurate first-fix estimation for reliable TDCP-based odometry. The findings contribute to highlighting the order of magnitude of errors on solutions as a function of the error on parameters
BiCoq : Bigraphs Formalisation with Coq
No full textInternational audienceBigraphs are a formal model for representing (ubiquitous) systems with strong notations of both space, e.g. a person in a room, and non-spatial relations, e.g. mobile phone communication regardless of location. They have been used in a wide range of scenarios including sensor systems, IoT configuration languages, and communications protocol design. While implementations of the bigraph theory exist, e.g. BigraphER, until now, there has been no attempt to formalise the theory in a theorem prover. We show an implementation of the bigraph theory in the Coq theorem prover, including the main bigraph type specification and common manipulation operators, e.g. composition and tensor product. This is a key step to fully formalising the theory and paves the way for a certified implementation for use in safety critical scenarios
Humanity does not play DICE
No full textThe DICE model claims that technological progress can prevent global warming at low cost. However, many uncertainties about climate damages and backstop technologies are not taken into account in this model. Our analysis is based on a very large number of scenarios to account for the possible range of the climate damage function and the level of efficiency of future backstop technologies, as well as their financial costs. We show that if DICE's assumptions turn out to be less optimistic, it's a very risky bet to rely on technology alone to tackle climate change, with 82% of our scenarios leading to economic collapse. Then we introduce in the DICE model a social lever called economic sufficiency, defined as the policy possibility to produce less than the maximum production capacity. The optimal DICE projections then choose economic sufficiency to mitigate climate damage or to avoid economic collapse
Firefighting with Drone Assistance: User Needs and Design Considerations for Thailand
No full textInternational audienceDrones are increasingly being deployed to assist firefighting crews in their missions, with the technology being chosen based on availability, rather than aligned with their specific needs. This phenomenon is exacerbated in the Global South, where infrastructure is scarce and where specific processes and user needs have to be adequately mapped to successfully introduce new technologies. We conducted semi-structured interviews with firefighting professionals (N=15) from Thailand, covering their prior experience with drones, challenges they encounter in their job, and how they envision this technology could better supportthem in the future. Our findings describe users’ technological needs and their expectations in terms of interaction and collaboration with drones. We identified specific challenges in Thailand that hinder the deployment of drone technology, including mismatches in technical and financial decisions. Furthermore, participants advocated for sharing physical systems between fire departments. We conclude with design considerations for drones in resource limited firefighting contexts
Vers des aides multimodales au pilotage en aviation légère : étude exploratoire par entretiens
No full textInternational audienceIn light aviation, perceptual errors represent a major problem for flight safety. To address this, we explore the use of multimodal alarm systems, which mobilise multiple senses. In order to improve our understanding of the possible challenges and opportunities associated with these systems, as well as of the perceptual and attentional deficits encountered by pilots, we conducted interviews with light aircraft pilots. The results describe an environment with a high visual and auditory load and highlight the difficulty of monitoring critical flight parameters in multitasking conditions. The interviewed pilots identified use cases where a deployment of multimodal interfaces could be beneficial. They also proposed types of alerts suitable to these use cases, such as gradual alarms or auditory and tactile feedback that allow monitoring of a changing parameter without needing to look at it. We conclude by discussing the implications for the design of multimodal systems tailored to light aviation.En aviation légère, les erreurs perceptives représentent un problème majeur pour la sécurité des vols. Pour y faire face, nous explorons l'usage d'alarmes multimodales, mobilisant plusieurs sens. Afin d'améliorer notre compréhension des possibles défis et des opportunités associés à ces systèmes, ainsi que des défaillances perceptives et attentionnelles rencontrées par les pilotes, nous avons mené des entretiens avec des pilotes d'aviation légère. Les résultats mettent en évidence un environnement déjà visuellement et auditivement chargé, où la surveillance de certains paramètres critiques devient difficile lors de tâches multiples. Les pilotes ont identifié des cas d'usage où un déploiement d'interfaces multimodales pourrait être pertinent et ont proposé des formes d'alerte adaptées, telles que des alarmes graduées ou des retours sonores et tactiles permettant de suivre l'évolution d'un paramètre sans le regarder. Nous concluons en discutant des implications pour la conception de systèmes multimodaux pour l'aviation légère
De la supervision au pilotage manuel : vers un contrôle flexible de flottes de drones pour la recherche et le sauvetage
No full textInternational audienceFleets of drones are actively being explored for search and rescue (SAR) missions to accelerate the exploration of large areas and support the delivery of supplies. However, the dynamic and unpredictable nature of these operations makes the automated planning and control difficult. We conducted interviews with four SAR professionals to identify critical scenarios and interaction needs at different levels of control. We then designed a virtual reality prototype allowing the planning, supervision, and manual takeover of a drone fleet during search missions. An exploratory study with five drone pilots evaluated the proposed interactions and helped identify design opportunities for such systems. Our results highlight the importance of enabling smooth manual takeover and the challenges of monitoring and controlling the fleet simultaneously. We discuss implications for the design of future SAR multi-drone systems.Les flottes de drones offrent un fort potentiel pour les missions de re- cherche et de sauvetage (SAR), en accélérant l’exploration de zones étendues ou le transport de matériel. Toutefois, la nature imprévi- sible et dynamique de ces interventions rend leur automatisation et leur supervision complexes. Nous avons mené des entretiens avec quatre professionnels du SAR pour identifier des scénarios critiques et des besoins d’interactions à différents niveaux de contrôle. Nous avons ensuite conçu un prototype en réalité virtuelle permettant de planifier, superviser et reprendre le contrôle manuel d’une flotte de drones lors de missions de recherche. Une étude exploratoire avec cinq télépilotes a permis d’évaluer les interactions proposées et d’identifier des opportunités de conception pour de tels systèmes. Nos résultats soulignent l’importance de permettre une reprise de contrôle fluide et les défis pour surveiller et contrôler la flotte en simultané. Nous discutons les implications pour la conception de futurs systèmes multi-drones
Une métrique rapide et précise pour évaluer et optimiser la complexité du trafic aérien.
No full textAir traffic actors seek to reduce air traffic's environmental impact. One future solution is to remove routes from the airspace so that aircraft can follow their optimal trajectories. These trajectories can consider both environmental and economic impacts. However, more complex traffic situations may then arise that prevent air traffic controllers from seeking safe and efficient operations. Air traffic management mitigates such issues and directly impacts the safety and efficiency of air traffic operations, passenger satisfaction, and airline objectives. It aims to maximize the use of airspace and minimize the impact on passengers, airlines, and the environment. The air traffic planning problem considers several factors to find a suitable solution, such as flight schedules, airspace capacity, flight routes, and environmental impact. Efficient solutions to this problem can improve air traffic management by reducing delays, diversions, and environmental impact. The airspace can be saturated before its capacity threshold due to the difficulty of air traffic controllers in managing air traffic. Therefore, a complexity metric is needed to assess their difficulty managing these various air traffic situations.There are several metrics for measuring the complexity of air traffic situations, but unfortunately, they do not capture every impact, such as uncertainty and difficulty in managing air traffic. For instance, the dynamic system metrics, while good measures of organizing a set of trajectories in a given airspace area, with uncertainty over time, do not fully account for the complexity of managing aircraft in a sector. Similarly, the solution state-based approach, while a good measure of the difficulty of managing aircraft in a sector using the permissible speeds and headings that would avoid other traffic, does not adequately consider the time uncertainty. Moreover, these metrics are very expensive in computation time, which is a disadvantage for optimizing the objective function. These limitations underscore the need for a new metric to comprehensively and efficiently measure air traffic complexity.This research aims to provide a metric that represents the difficulty of managing air traffic situations and is quick to calculate. The proposed metric estimates the total conflict duration between two aircraft for heading change maneuvers (or climbing rate uncertainties) and probable passage times. The proposed metric considers the time uncertainty, which results in several possible aircraft positions along the trajectory at a given time. For each trajectory, these possible positions form the set of observations at a given moment. For a pair of observations in the set defined above, the proposed metric numerically integrates the duration of conflict for each heading change if the aircraft is cruising or for each climbing rate uncertainty if the aircraft is climbing/descending. The metric then sums the total conflict duration for each instant to give a metric of the difficulty of all air traffic. The metric is validated using distinct cases of air traffic situations to discriminate their level of difficulty in managing.This research addresses a complex and challenging problem in air traffic management, one that requires advanced black-box optimization algorithms and tools. The size of the problem encourages the use of metaheuristics. Solving this optimization problem with a dedicated simulated annealing seems promising. As a result, algorithms are likely to require numerous and costly evaluations of the objective function. This problem underscores the need for innovative solutions, such as the metric proposed in this research, to define the complexity/difficulty of air traffic situations with a low computation time. The simulated data of French airspace traffic with over 8,000 flights validates the combination of the metric with a dedicated simulated annealing algorithm.Les acteurs du trafic aérien cherchent à réduire son impact environnemental. Une solution consiste à supprimer les routes fixes pour permettre aux avions de suivre leurs trajectoires optimales, tenant compte des impacts environnementaux et économiques. Cependant, des situations de trafic plus complexes peuvent surgir, rendant la tâche des contrôleurs aériens plus difficile. La gestion du trafic aérien atténue ces problèmes et influence directement la sécurité, l'efficacité des opérations, la satisfaction des passagers et les objectifs des compagnies aériennes. Elle vise à optimiser l'utilisation de l'espace aérien tout en minimisant les impacts sur les passagers, les compagnies et l'environnement. La planification du trafic aérien doit considérer les horaires de vol, la capacité de l'espace aérien, les routes de vol et l'impact environnemental. Des solutions efficaces peuvent réduire les retards, les détournements et l'impact environnemental. L'espace aérien peut être saturé avant d'atteindre sa capacité maximale en raison de la complexité de gestion pour les contrôleurs aériens. Une métrique de complexité est donc nécessaire pour évaluer cette difficulté.Il existe plusieurs métriques pour mesurer la complexité du trafic aérien, mais elles ne capturent pas tous les impacts, comme l'incertitude et la difficulté de gestion. Par exemple, les métriques de système dynamique et l'approche basée sur l'état de la solution ne tiennent pas pleinement compte de la complexité et de l'incertitude temporelle de la gestion des avions. De plus, elles sont coûteuses en temps de calcul, ce qui est un inconvénient pour l'optimisation. Ces limitations montrent la nécessité d'une nouvelle métrique pour mesurer efficacement et exhaustivement la complexité du trafic aérien.Cette recherche vise à fournir une métrique qui représente la difficulté de gestion des situations de trafic aérien et qui est rapide à calculer. La métrique proposée estime la durée totale de conflit entre deux avions pour des manœuvres de changement de cap (ou des incertitudes de taux de montée) et des temps de passage probables. La métrique proposée prend en compte l'incertitude temporelle, ce qui entraîne plusieurs positions possibles des avions le long de la trajectoire à un moment donné. Pour chaque trajectoire, ces positions possibles forment l'ensemble des observations à un moment donné. Pour une paire d'observations dans l'ensemble défini ci-dessus, la métrique proposée intègre numériquement la durée du conflit pour chaque changement de cap si l'avion est en croisière ou pour chaque incertitude de taux de montée si l'avion monte/descend. La métrique additionne ensuite la durée totale du conflit pour chaque instant pour donner une métrique de la difficulté de tout le trafic aérien. La métrique est validée à l'aide de différents cas de situations de trafic aérien pour discriminer leur niveau de difficulté de gestion.Cette recherche aborde un problème complexe et difficile dans la gestion du trafic aérien, nécessitant des algorithmes et des outils avancés d'optimisation de type ''boîte noire''. La taille du problème encourage l'utilisation de métaheuristiques. La résolution de ce problème d'optimisation avec un recuit simulé dédié semble prometteuse. En conséquence, les algorithmes sont susceptibles de nécessiter de nombreuses évaluations coûteuses de la fonction objective. Ce problème souligne la nécessité de solutions innovantes, telles que la métrique proposée dans cette recherche, pour définir la complexité/difficulté des situations de trafic aérien avec un temps de calcul réduit. Les données simulées du trafic de l'espace aérien français avec plus de 8 000 vols valident la combinaison de la métrique avec un algorithme de recuit simulé dédié
Varif.ai to Vary and Verify User-Driven Diversity in Scalable Image Generation
No full textInternational audienceFigure 1: Varif.ai supports user-defined diversity to S1) generate images, S2) verify their diversity based on attributes, and S3) vary diversity iteratively. Left: Images generated with a state-of-the-art diffusion model [40] using the prompt "a picture of a car" lacks diversity. Right: Images generated with Varif.ai has more diversity in car color, car model, and background landscape.</div