Aircraft State Research Articles

Human Factors Approaches and Models in LOC-I Accident Analysis and Prevention: Flight Crew Resource Management Techniques as a Risk Mitigation Tool

Increased cockpit automation on modern jet aircraft aim to reduce the risk of Undesired Aircraft State (UAS) instances such as Loss of Control in Flight (LOC-I). Although LOCI globally accounts for only 9% of all analysed accidents IATA has reported that it was responsible for 58% of all accident fatalities in 2017. The focus of this paper is to answer whether Threat and Error Management and Crew Resources Management (CRM) techniques are an efficient risk management tool when facing a LOC-I threat. Three LOCI final aircraft accident reports were analysed to understand the structure of Human Factors (HF) during these flights. Methods from the HF field such as the Generic Error Modelling System (GEMS) and Skill-, Rule- and Knowledge-based (SRK) error approach provided invaluable insights to identify potential findings. A holistic investigation of cognitive structures in flight path management helped to visualise latent conditions and cognitively demanding tasks during LOC-I in routine operations. Bearing in mind the limited number of cases considered in this paper it should be considered as an overview in LOC-I accident analysis. It shows that leadership and teamwork, as essential aspects of CRM training, can serve as key strategies to mitigate HF problems and LOC-I risks.

Space eye on flying aircraft: From Sentinel-2 MSI parallax to hybrid computing

Knowledge of the status (position and speed) of flying aircraft is vital for efficient and safe air space management. However, this requirement is often compromised due to the complexity of the aviation environment. Satellite remote sensing (RS) provides a complementary means for tracing aircraft, but is often limited to finding motionless aircraft under specific scenarios (e.g., at airports). Here, based on the inter-band offsets due to hardware parallax in push-broom sensors (e.g., Sentinel-2 Multispectral Instrument or MSI), we develop a method for detecting flying aircraft in an automated fashion. Supported by a hybrid computation framework (based on Google Earth Engine computation and local computation) specifically designed to address the challenge of processing large volume of moderate resolution RS data at a global scale, the method is applied to more than 2.31 million MSI images to establish a map of the global distribution of flying aircraft. The detected flying aircraft coincide well with those determined using traditional techniques (e.g., Flightradar24), when both datasets co-exist. With the existing and future moderate-resolution data captured by push-broom satellite sensors, the method is believed to provide a robust and cost-effective means of detecting aircraft status at a global scale, thus supplementing the traditional methods for tracking flying aircraft. The same method is also used to estimate the inter-band and inter-granule time offsets in multi-band MSI and Landsat-8 Operational Land Imager (OLI) images, which may provide critical information needed to correct artifacts in aquatic applications.

Autolanding control system design with deep learning based fault estimation

Developing a control system that can recover the aircraft under actuator failures and severe disturbances is a popular problem in flight control system design. Majority of existing controllers cannot adapt to changes in aircraft dynamics that occur due to severe actuator failures. We propose a novel, data-driven fault estimation method for estimating actuator faults from aircraft state trajectories, which utilizes a deep neural network trained offline on gathered simulation data of fault injected aircraft. Proposed novel deep fault estimation model coupled with an existing nonlinear dynamic inversion based autolanding controller, reacts immediately to a wide range of actuator failures and is able to land the aircraft under many different combinations of actuator failures and severe wind conditions. Performance of the proposed approach is compared to existing state-of-the-art fault tolerant controllers through fault tolerance maps. It is observed that the developed approach is superior both in terms of fault tolerance map coverage and smoothness of the computed controller signals.

A Vision-Based Method for Determining Aircraft State during Spin Recovery.

This article proposes a vision-based method of determining in which of the three states, defined in the spin recovery process, is an aircraft. The correct identification of this state is necessary to make the right decisions during the spin recovery maneuver. The proposed solution employs a keypoints displacements analysis in consecutive frames taken from the on-board camera. The idea of voting on the temporary location of the rotation axis and dominant displacement direction was used. The decision about the state is made based on a proposed set of rules employing the histogram spread measure. To validate the method, experiments on flight simulator videos, recorded at varying altitudes and in different lighting, background, and visibility conditions, were carried out. For the selected conditions, the first flight tests were also performed. Qualitative and quantitative assessments were conducted using a multimedia data annotation tool and the Jaccard index, respectively. The proposed approach could be the basis for creating a solution supporting the pilot in the process of aircraft spin recovery and, in the future, the development of an autonomous method.

Identification of Key Flight Conflict Nodes Based on Complex Network Theory

The air traffic density in the terminal area is high and the traffic situation is relatively complex by the development of aviation, which brings great challenges to controller. In order to understand the flight situation and provide decision basis for controllers, this paper proposes a key flight conflict nodes identification method based on complex network theory and Analytic Hierarchy Process (AHP)-entropy weight method. Firstly, an aircraft state network is established with aircraft as nodes and Airborne Collision Avoidance System (ACAS) communication relations as connecting edges. On this basis, four parameters, node degree, node weight, clustering coefficient and betweenness, are selected as evaluation indexes of node importance, and the weight of each index is determined by using AHP. And entropy weight method is introduced to revise the results. Node importance is calculated through multi-attribute decision-making method to determine key conflict aircrafts. The simulation and experiment on the artificial network and the aircraft state network of a certain day in the terminal area of Kunming Changshui Airport show that the method proposed in this paper can identify the key flight conflict nodes in the aircraft state network, allocate the selected node deployment can effectively reduce the complexity of the aircraft state network, can provide reference for air traffic control services (ATCS), and reduce the allocation difficulty of controller.

Using Broadcast Wind Observations to Update the Optimal Descent Trajectory in Real-Time

The ability to meet a controlled time of arrival during a continuous descent operation will enable environmentally friendly and fuel-efficient descent operations while simultaneously maintaining airport throughput. However, if the wind forecast used to compute the initial trajectory plan is not accurate enough, the guidance system will need to correct time deviations from the plan during the execution of the descent. Previous work proposed an onboard guidance strategy based on model predictive control, which repeatedly updates the trajectory plan in real-time from the current aircraft state and for the remainder of the descent. However, the wind conditions downstream, at altitudes not explored yet, were difficult to predict due to the lack of data. This paper shows the potential benefits of using wind observations, broadcast by nearby aircraft, to reconstruct the wind profile downstream. The wind profile in the trajectory optimization problem is modeled as a spline, control points of which are updated to fit the observations before replanning the trajectory. Results from simulations using realistic wind data show that the performance of model predictive control significantly improves when including up-to-date wind observations, in terms of time and energy errors at the metering fix and fuel consumption.

Performance-Based Ice Detection Methodology

A novel robust ice detection methodology for the early detection of icing-related flight performance degradation is presented. Based on data of 75,689 flights with modern commercial airliners, a maximum aircraft fleet’s performance variation has been estimated. The evaluation of results indicates that an expected influence of icing could be clearly separated. The developed methodology is energy based and fuses aircraft body and engine influences on flight performance, which allows to reliably calculate a deviation from an available reference. This difference in flight performance is consequently used to detect an aerodynamic degradation. The novel methodology provides large capabilities and shows a good detection reliability with no false alarm even within maneuvering flight, wind shear, turbulence, and sideslip as well as for several sensor error cases. The methodology was evaluated during various simulator trials. The results show a very high potential in supporting pilots with adequate information about the current aircraft status in icing conditions.

Interlayer Working Conditions Classification and Treatment Measures of Airport Asphalt Pavement Overlay

The bonding strength between the overlays of an airport asphalt pavement directly affects its service life. In this study, a finite element model of asphalt overlay on airport pavement of Juba Airport in South Sudan was established, and the elastic modulus of overlay, thickness of overlay, running state of aircraft, variation trend of interlayer shear stress under the action of load and temperature were obtained. The AHP-entropy method was used to calculate the combined weight of each influencing factor. The climate in hot and humid areas was further zoned, and a comprehensive classification of airport asphalt pavement overlay between hot and humid tropics was obtained. Four interlayer treatment measures (asphalt precoating, two oils and one aggregate, geotextile, and geogrid) and three interlayer materials (SBS-modified asphalt, ordinary hot asphalt, and SBS-modified emulsified asphalt) were designed and tested. Through shear test, wheel-load fatigue test, and pressurized seepage test, the variation trends of interlayer shear, crack resistance, and waterproofing under different combination schemes were obtained. Finally, the matching relationship between different working conditions and interlayer treatment measures was established. The results show that the actual working conditions of an airport asphalt pavement can be divided into three levels. Combined with the performance ranking of different materials and interlayer treatment measures, the recommended interlayer treatment measures for comprehensive classifications I, II, and III are two oils and one material + SBS-modified asphalt, asphalt precoating, or geogrid + SBS-modified emulsified asphalt or ordinary hot asphalt, and geotextile + ordinary hot asphalt, respectively. The working conditions in Juba Airport in South Sudan can be classified as comprehensive classification II, the recommended treatment measures for geogrid + ordinary hot asphalt.

Development of Smart Aircraft Logbook Based on IoT

This project is to develop a system that can integrate between user input and data reference stored in a cloud database. The user input is referring to the post-flight report regarding aircraft status retrieved from the Flight Management System (FMS). The data will be crosschecked subject to the reference manual which is the Master Minimum Equipment List (MMEL) and o Minimum Equipment List (MEL) that have been kept in a cloud database. This system is constructed to reduce time and response time for the post-flight check. Research exploratory had been carried out using the interview method, to obtain the information regarding the application and implementation of the Internet of Things (IoT) in the aviation industry. System reference is extracted into the cloud database. Next, the software is developed so that the user can have interaction with this system. From all the fragments from the input and data reference, this system is recommended a fast solution to determine the aircraft status; whether it is safe to fly or to be grounded for maintenance purposes. Through this methodology and IoT concept, it is believed that this system may be a beneficial idea and effort to fit the aviation industry's needs towards the rise of the Industrial Revolution 4.0.

A model of interior cockpit noise for the F-35A

Characterization of environmental noise is vital to protection of hearing. A measurement campaign to understand the nature of the noise in the cockpit of fifth-generation fighter aircraft has been completed. Measurements during multiple missions for each variant of the F-35 aircraft characterized the cockpit acoustic levels during various stages of flight. Additional data collection defined the aircraft state during the acoustic data collection. An initial model of the F-35A represents the A-weighted overall sound pressure level (LA) within the cockpit by a series of random decision forest models. Sensitivity analysis conducted on these models demonstrates that airspeed and the environmental control system (ECS) flow rate are the two most important parameters determining the interior acoustic levels. However, the measurements cover a subset of the full dynamic envelope of the aircraft. Since machine-learning algorithms are susceptible to greater uncertainty where no acoustic data exists, a curve fit model defines acoustic levels throughout the envelope. This model adds the ECS flow rate to the dynamic pressure calculations and produces a more accurate model of LA within the cockpit of the F-35A. This research is sponsored, in part, by F-35 Joint Program Office.

Incremental-Learning-Based Unsupervised Anomaly Detection Algorithm for Terminal Airspace Operations

The terminal airspace system is one of the most complex human-made dynamical systems with a continuously evolving structure. Modern aviation technologies in the terminal airspace record detailed flight and airport information, including the states of aircraft and the operating conditions of airports as time series datasets, through on-board and ground-based systems. Detecting anomalies in such datasets is emerging as a key problem to understand air transportation system complexity and behavior. This paper proposes an incremental-learning-based anomaly detection algorithm that generates anomaly detection models from surveillance data. The algorithm is developed in a recursive fashion, such that it uses the overnight recorded data to improve the temporal-logic-based anomaly detection models daily—thus it is termed as an overnight update algorithm. The proposed algorithm is demonstrated with real air traffic surveillance data comprising of arrival flights to LaGuardia, John F. Kennedy, and Newark airports, thereby detecting anomalies for arrivals in the terminal airspace of the New York metroplex. This is followed by some analysis and insights from the results of these extensive tests.

Pilot competencies as components of a dynamic human‐machine system

AbstractThis paper formulates a theoretical model of flight deck team performance as a dynamic human‐machine system, later referred to as CRM‐DYMO (Crew Resources Management‐DYnamic MOdel). The model identifies human and machine subsystems. The human subsystem describes the activities taken by the flight crew and the machine subsystem describes the changes in the aircraft's state. The human subsystem's output forms an input for the machine subsystem, and the machine subsystem's output feeds back to the human system creating a closed loop system. Pilots’ performance are described as competencies, which form part of the elements in the human subsystem. This paper concentrates only on CRM competencies as defined by the International Civil Aviation Authority and the International Air Transport Association. In addition to the detailed conceptual model of the human subsystem and associated competencies, the approach is illustrated with reference to two accident case studies. These demonstrate how CRM‐DYMO can assist in analyzing CRM and help in identifying how the different system components interact. Although the case studies demonstrate the benefits of CRM‐DYMO, they merely serve as a starting point to understand how CRM works and will hopefully encourage supporting quantitative simulation modeling and empirical studies.

3-D tracking of air targets using a single 2-D radar

Three-dimensional (3-D) aircraft tracking using two-dimensional (2-D) radar measurements is a common task in air traffic control (ATC) systems. In this paper, a new algorithm using a 2-D ATC radar, called HPEKF with probabilistic data association (HP-PDA-EKF), is proposed for aircraft tracking with false alarms and missed detections. An improved filter initialization technique is also proposed to yield better estimates. To mitigate the degradation in altitude estimation accuracy with increasing aircraft distance, ATC waypoints are used as extra measurement information, and a new height-parametrized cascaded PDA filter (HP-CPDAF) algorithm is proposed to make use of waypoints. The posterior Cramér–Rao lower bounds (PCRLBs) quantifying the achievable accuracies of these two filters are derived. Simulations are carried out to analyze the relationship between aircraft distance, waypoint accuracy and state estimation accuracy. The viability of using a single 2-D radar for 3-D ATC and the usefulness of the proposed algorithms under different operating conditions are analyzed. Simulation results demonstrate the validity of the proposed algorithms.

Implementing Augmented Reality in the Flight Deck for Single Pilot Operations

Single Pilot Operations is a current topic with the potential to significantly affect the future of commercial aviation. While financially attractive for airlines, Single Pilot Operations bring forth important safety concerns, especially regarding the lack of human redundancy in the flight deck, an increased workload for the single pilot, reduced situational awareness and a higher risk of human error.
 It is assumed that potential problems affecting Single Pilot Operations could be addressed by implementing an Augmented Reality (AR) device in the flight deck, by presenting additional information and supporting hints within the pilot’s field of view. Concretely, AR could be used to help reduce the single pilot’s workload, improve situational awareness and reduce the risk of human error.
 This paper sets out to demonstrate two use cases for augmented reality in the flight deck. A system, called Pilot Assist, was developed that allows pilots to conduct checklists interactively with a Microsoft HoloLens. The system also provides a holographic Head-up-Display. Pilot Assist was developed and demonstrated with a fixed base Airbus A320 simulator at the Technical University of Wildau.
 With the HoloLens’ spatial mapping capabilities – scanning and recognizing the environment around the user – it was possible to create a system that guides the pilot through the conduction of checklists. This is done by prompting the user towards the location of each checklist item in the cockpit, where information regarding necessary actions is projected. Furthermore, Pilot Assist is integrated with the aircraft systems, making it possible to obtain aircraft status data in real time, thus allowing error-checking of the pilot’s actions as well as automating the progress through checklists.
 The holographic Head-up-Display allows the user to look at the surrounding environment while presenting critical flight data within the user’s field of view. The holographic Head-up-Display is intended to contribute to the pilot’s situational awareness.
 Experts in the aviation field, including pilots, researchers and engineers had the chance to qualitatively assess the Pilot Assist tool. They pointed to limitations of both Pilot Assist and the HoloLens itself, but shared optimism as to how this technology and similar applications could indeed impact the future of flight operations. Concerns regarding the HoloLens’ weight, comfort and narrow field of view were expressed. However, continued development of head mounted devices (e.g. HoloLens 2) is expected in the coming years.
 Further research into augmented reality applications in the flight deck is needed to advance this and other use cases. Nonetheless, the experts agreed Pilot Assist provides beneficial support during single pilot operation considering the current prototypical nature of the system.

PyModeS: Decoding Mode-S Surveillance Data for Open Air Transportation Research

The availability of low-cost automatic dependent surveillance-broadcast (ADS-B) receivers has given researchers the ability to make use of large amounts of aircraft state data. This data is being used to support air transportation research in performance study, trajectory prediction, procedure analysis, and airspace design. However, aircraft states contained in ADS-B messages are limited. More performance parameters are downlinked as Mode-S Comm-B replies, upon the automatic and periodic interrogation of air traffic control secondary surveillance radar. These replies reveal aircraft airspeed, turn rate, target altitude, and so on. They can be intercepted using the same 1090-MHz receiver that receives the ADS-B messages. However, a third-party observer does not know the interrogations, which originated the Comm-B replies. Thus, it is difficult to decode these messages without knowing the type and source aircraft. Furthermore, the parity check also cannot be performed without knowing the interrogations. In this paper, we propose a new heuristic-probabilistic method to decode the Comm-B replies and to check the correctness of the messages. Based on a reference dataset provided by air traffic control of the Netherlands, the method yields a success rate of 97.68% with an error below 0.01%. The performance of the proposed method is further examined with data from eight different regions of the world. The implementation of the inference and decoding process, pyModeS , is shared as an open-source library.

Aircraft Surface Trajectory Prediction Method Based on LSTM with Attenuated Memory Window

With the rapid development of the global air transport industry, airport surface traffic is increasingly busy. The safety hazards of taxiing conflict still exist during the airport operation for the initially planned taxiing path about aircraft, which directly affects the operational safety and efficiency of the airport surface. Using the correlation and dependence between the position sequences of aircraft glide motion, a method based on Long Short Term Memory networks (LSTM) is used. Combining the change of the motion state of the surface aircraft, the attenuation memory window is introduced to improve the hidden layer structure to further enhance the prediction accuracy in the LSTM model, and it is compared and verified under different parameters. The method realizes the target of surface position prediction for the aircraft in future period, lays the foundation for aircraft surface taxiing conflict detecting, and avoids taxiing conflict.

Comparison of Conflict Detection Capabilities for Well Clear and Critical Pair Identification

As air traffic management becomes increasingly automated, technologies for detecting and resolving conflicts are critical to maintain the safety of an airspace. There exists a need to understand how multiple conflict detection models may work together to provide deeper insights into the safety of an airspace with many aircraft interactions. Conflict detection capabilities can be improved by using two recently introduced models, Well Clear (WC) and Critical Pair Identification (CPI), which address separation assurance and detection of near mid-air collisions due to unexpected maneuvers, respectively. A procedure called Pairwise Sweep of a Pairwise Conflict (PSPC) is provided, which allows for the visualization of the response of WC and CPI for conflicts with a variety of initial aircraft states in the horizontal airspace. PSPC is then used to analyze a two-aircraft scenario using both WC and CPI to create a merged response. Their joint capability can benefit unmanned aerial systems as a conflict detection technology and air traffic controllers as a decision support tool.

Developing Active Manipulators in Aircraft Flight Control

Most haptic interfaces developed for aircraft control provide haptic support as an additional force on the control manipulator. This study revisits the active manipulator, which is a design concept that is different from but complementary to existing haptic interfaces. This control device sends the force that the pilot exerts on it to the aircraft while feeding back the aircraft rotational rate by means of its deflection angle. It is found that, in comparison with the conventional passive manipulator, the active manipulator greatly facilitates target following and disturbance rejection in compensatory tracking tasks. Furthermore, larger improvements in task performance are associated with higher forcing-function bandwidths. These findings are accounted for by the fact that the active manipulator changes the effective controlled-element dynamics into integratorlike dynamics while at the same time integrating disturbance rejection into the neuromuscular system. However, the high-frequency disturbances acting on the aircraft present in feedback about the aircraft state adversely affect the operational effectiveness of the active manipulator. Based on the experimental findings and results from the passivity theory, a lead–lag filter is designed and evaluated, which mitigates this effect without affecting task performance.

POLISH EXPERIENCE IN OPERATIONAL AIR TRAFFIC MANAGEMENT IN THE CONTEXT OF EUROPEAN AIRSPACE INTEGRATION

The continued air traffic growth in Polish airspace requires continuous improvement of the efficiency of the air traffic management (ATM) system in order to meet the needs of all its users and to guarantee the highest level of safety of air operations. Activities in this respect undertaken by the Polish Air Navigation Services Agency (PANSA) in cooperation with military aviation authorities are in line with the objectives of the EU Single European Sky (SES) initiative. In addition to the provision of air navigation services (ANS) for civil aviation (commercial air transport and general aviation), PANSA also provides navigation services for the crews of military aircraft and other state aircraft flying in operational air traffic (OAT). The article describes in general terms flights in operational air traffic, presents the evolution of European regulations concerning this category of air traffic and discusses in detail selected procedures for planning and protecting military aviation flights in Polish airspace.

Infrared-Based Autonomous Navigation for Civil Aircraft Precision Approach and Landing

Precise navigation is a fundamental problem of aircraft safety approach and landing. However, the existing methods, including rotorcraft-based and fixed-wing-based, cannot meet the requirements of precision approach and landing of civil aircraft in the global position system (GPS)-denied and low visibility. This paper proposes an autonomous approach and landing navigation method whose accuracy is comparable to Inertial/Differential GPS (DGPS) integration. This method integrates inertial data, forward-looking infrared (FLIR) images, and runway geographic information to estimate kinetics states of aircraft during approach and landing. First, we improve an existing method to robustly detect runway, accurately extract three vertexes of runway contour from FLIR images, and synthesize the virtual runway features by runway geo-information and aircraft's pose parameters. Second, we propose to use real and synthetic runway features to create vision cues and integrate them with inertial data in square-root unscented Kalman filter to estimate the motion errors. Meanwhile, the measured motion states are corrected with the estimated state errors. Finally, we design a flight data acquisition platform equipped on a general aircraft and use the real flight data to verify our proposed method. The experimental results demonstrate that the proposed method can run smoothly for civil aircraft precision approach and landing.