Archive ouverte de l'ENAC
Not a member yet
3458 research outputs found
Sort by
A New Geometric Regression with Inputs-Outputs on Matrix Lie Groups
No full textInternational audienceThis paper investigates a new Lie group regression model for input-output data belonging to Lie groups. The originality of the model lies in the fact that the unknown weights also lie in Lie groups and are learned using an intrinsic optimization algorithm based on maximum likelihood estimation. The model is validated through numerical simulations conducted using synthetic data belonging to the Lie group SO(3), which is commonly used in robotics to represent rotational observations
FLIGHT ALLOCATION TO EN-ROUTE CONTROLLERS IN THE TBO FRAMEWORK
No full textInternational audienceThe present communication deals within the Trajectory Based Operations framework, on the problem of assigning flights to controllers in extended control areas. First, a flight centric complexity metric is used to quantify the interaction between planned flights at the pre-tactical level and an assignment problem solved through an heuristic while leads to the dimensionning of the control task force. Then, to cope with the tactical level, the flight centric complexity metric is modified to take into account trajectories' uncertainty. This allows to propose an on-line allocation scheme of flights to controllers according to their workload and to the estimated traffic complexity expected to be met by these flights
Flight allocation in flight-centric air traffic control: A MILP model approach
No full textInternational audienceThis study addresses flight allocation in a flightcentric air traffic management framework using two Mixed-Integer Linear Programming (MILP) models. The first model minimizes maximum complexity to balance workload, while the second minimizes interactions between flights assigned to different controllers. Tests over the Brest airspace show that the first model, despite achieving workload balance, leads to impractical allocations by assigning distant, non-interacting flights to the same controller. The second model, though more operationally feasible, results in workload imbalances, with some controllers managing high-complexity traffic while others remain underloaded. Despite this limitation, it better preserves spatial coherence and allows for fast adjustments to traffic changes. Future work should integrate conflict detection and resolution to optimize allocations while considering real-time controller actions.</div
SafeSpect: Safety-First Augmented Reality Heads-up Display for Drone Inspections
No full textInternational audienceDrones are increasingly being deployed to assist firefighting crews in their missions, with the technology being chosen based onavailability, rather than aligned with their specific needs. This phenomenon is exacerbated in the Global South, where infrastructure isscarce and where specific processes and user needs have to be adequately mapped to successfully introduce new technologies. Weconducted semi-structured interviews with firefighting professionals (N=15) from Thailand, covering their prior experience withdrones, challenges they encounter in their job, and how they envision this technology could better support them in the future. Ourfindings describe users’ technological needs and their expectations in terms of interaction and collaboration with drones. We identifiedspecific challenges in Thailand that hinder the deployment of drone technology, including mismatches in technical and financialdecisions. Furthermore, participants advocated for sharing physical systems between fire departments. We conclude with designconsiderations for drones in resource-limited firefighting contexts
Towards a new approach to ensure end-to-end reliability of aeronautical data communications
No full textInternational audienceAbstract The aeronautical telecommunication network (ATN) aims to provide reliable end-to-end communications even for those including the air-to-ground segment and in particular for data link applications. The existing ATN, known as ATN/OSI, is based on OSI protocols since its first deployment. The OSI model implementation in ATN communicating entities causes great complexity in network management, particularly in terms of Internet network interoperability. Therefore, since 2010, the International Civil Aviation Organization (ICAO) proposed a migration to ATN over Internet protocol suite (IPS), called ATN/IPS. Thus, this research work focuses on specifying the reliability mechanisms required for air ground data link applications in future ATN/IPS. To achieve this, the transport protocols performance is assessed based on simulations using an ATN model developed considering the ICAO standards. The modeled legacy application enables to generate traffic based on real controller-pilot data link communications (CPDLC) log files from French area control centre (ACC). The air-to-ground subnetworks are characterised using time series delay induced from previously modeled VDL Mode 2 data link analysis. As proof-of-concept, CPDLC messages exchange from aircraft to controller and future applications that transmits heavier files from ground-to-board are simulated. Transport protocols performance are evaluated with respect to the most constraining requirements. The simulation results highlighted the limitations of both connection-oriented transport protocol class 4 (COTP4) and TCP. This enabled to provide a preliminary overview of a new QUIC-like reliable protocol that should meet the heterogeneous requirements of the legacy and the eventual future ATN/IPS applications
Formally proved specification of non-nested STL formulas as synchronous observers
No full textInternational audienceSignal Temporal Logic (STL) is a convenient formalism to express bounded horizon properties of autonomous critical systems. STL allows to express real-valued signal properties and associates a non-singleton bound interval to each temporal operators. In the case of critical autonomous systems, it may be necessary to check the validity of an STL property in real-time. To that end, we provide a rigorous encoding of non-nested discrete-time STL formulas into Lustre synchronous observers.Our encoding provides a three-valued online semantics for the observers and therefore enables both the verification of the property and the search of counter-examples. A key contribution of this work is an instrumented proof of the validity of the implementation with respect to the original STL semantics. All of the experiments are automated with the Kind2 model checker and the Z3 SMT solver
Numerical optimization of aviation decarbonization scenarios: balancing traffic and emissions with maturing energy carriers and aircraft technology
No full textDespite being considered a hard-to-abate sector, aviation’s emissions will play an important role in long-term climate mitigation of transportation. The introduction of low-carbon energy carriers and the deployment of new aircraft in the current fleet are modeled as technology-centered decarbonization policies, while supply constraints in targeted market segments are modeled as demand-side policies. Shared Socioeconomic Pathways (SSPs) are used to estimate trend-mitigation traffic demand and to limit the sectoral consumption of electricity and biomass. Mitigation scenarios are formulated as optimization problems, and three applications are demonstrated: no-policy baselines, single-policy optimization, and scenario-robust policies. Results show that the choice of energy carrier is highly dependent on assumptions regarding aircraft technology and the background energy system. Across all SSP-based scenarios, emissions peak by around 2040, but achieving alignment with the Paris Agreement requires either targeted demand management or additional low-carbon energy supply. The use of gradient-based optimization within a multidisciplinary framework enables the efficient resolution of these nonlinear, high-dimensional problems while reducing implementation effort
Interpretable and Explainable Surrogate Modeling for Simulations: A State-of-the-Art Survey and Perspectives on Explainable AI for Decision-Making
No full textSurrogate models play a central role in reducing the computational cost of complex systems simulations across a wide range of scientific and engineering domains. Yet, their black-box nature often hinders insight into how input variables drive system responses. This state-of-the-art review and position paper surveys the integration of surrogate modeling with explainable artificial intelligence techniques to enhance transparency in simulation workflows and support design and decision-making, while also offering perspectives on future research directions. The paper not only surveys a broad spectrum of explainability techniques, such as variance-based sensitivity analysis, partial dependence plots, SHAP value decomposition, and active subspace methods, but also discusses practical pathways for applying them, including integration into design exploration, multi-objective design, and decision-making processes. We highlight their strengths for revealing interactions, managing high-dimensional and correlated inputs, and supporting human comprehension. This combination of methodological coverage and practical guidance uniquely positions the paper at the intersection of method development and practical application in simulation workflows. Beyond surveying existing methods, we also identify pressing challenges, including the explainability of dynamical systems, mixed-variable systems, correlated input structures, and the robustness of explanation metrics. We conclude with a research agenda to make interpretability and explainability core elements of simulation-driven workflows, embedded from model construction through validation and sensitive to context, modeling choices, and the limits of simulation for design and decision support
Réexaminer les surprises liées à l'automatisation : une étude centrée sur les défauts de conception du pilote automatique
No full textThis thesis investigates how specific design choices in aircraft autopilot systems contribute to automation surprise—when the aircraft behaves differently from the pilot's expectations. It focuses on two key interfaces: the Flight Mode Annunciator (FMA) and the Flight Control Unit (FCU). The FMA displays autopilot modes in use and has remained unchanged since the 1970s, with most manufacturers adopting the same design despite longstanding concerns raised in the literature. In contrast, the FCU, which allows pilots to select these modes, varies across aircraft manufacturers in both layout and interaction design. While the FMA is difficult to interpret and monitor, the FCU raises concerns because of overloaded controls. Across both interfaces, the industry has historically relied on training and procedural workarounds to address design flaws rather than rethinking interaction design. Grounded in Human-Computer Interaction (HCI) theories, this thesis adopts a mixed-methods, primarily qualitative approach, combining empirical studies with pilots and a participatory design process to explore the roots of automation surprise and propose design alternatives to better support mode awareness. This thesis contributes to ongoing research on automation in safety-critical systems and argues for a shift toward co-designing cockpit interfaces with pilots, rather than merely for them.Cette thèse examine comment certains choix de conception dans le système de pilotage automatique des avions contribuent aux surprises liées à l'automatisation—lorsque l'avion se comporte différemment de ce que le pilote attend. Elle se concentre sur deux interfaces clés : le Flight Mode Annunciator (FMA) qui affiche les modes utilisés par le pilote automatique, et le Flight Control Unit (FCU) qui permet aux pilotes de les sélectionner. Depuis les années 1970, le design du FMA est resté largement inchangé, la plupart des avionneurs continuant à utiliser un format fréquemment critiqué dans la littérature comme étant difficile à interpréter et à surveiller. Le FCU, quant à lui, varie selon les constructeurs, tant dans sa disposition que dans son design d'interaction et soulève des préoccupations liées aux types de boutons. Pour pallier aux défauts de conception de ces deux interfaces, l'industrie s'est historiquement appuyée sur la formation et les procédures plutôt que de repenser l'interaction elle-même. Ancrée dans les théories de l'Interaction Humain-Machine (IHM), cette thèse adopte une approche mixte, majoritairement qualitative, combinant des études empiriques avec des pilotes et une démarche de conception participative pour explorer les origines des surprises liées à l'automatisation et pour proposer des alternatives de design permettant de mieux soutenir la conscience des modes. Cette dissertation contribue aux travaux en cours sur l'automatisation dans les systèmes critique et soutient l'idée que la conception des interfaces de cockpit ne devrait pas se limiter à des solutions imposées aux pilotes, mais s'inscrire dans une démarche collaborative avec eux
Spin Spherical Harmonics for the Analysis of Antenna Electromagnetic Fields
No full textInternational audienceSpherical harmonics are classical analysis tools in many science and engineering domains. For analyzing the electromagnetic fields of antennas in the frequency domain, the mostly used formulation is the one proposed by Hansen. This article proposes an alternative solution, relying on spin spherical harmonics. On a sphere, the tangential components of the electric and magnetic fields are represented by means of harmonics of spin ±1. Then new closed-form relations are established between the spin spherical harmonics and the ones formulated by Hansen. A sampling theorem and fast transforms that are consistent with spin spherical harmonics are used. The radiations of spin spherical harmonics of order 1 are related to elementary dipoles and Huygens sources in circular polarization. Finally, numerical experiments are performed with a horn antenna and a GNSS antenna installed on an aircraft. They show that a very large radiating system with a band-limit of 2048 can be efficiently analyzed by means of fast spin spherical harmonic transforms, with a computation time of 2 minutes, approximately