Search

This paper presents a new challenge on the strategic 4D trajectory optimization problem with the evaluation of air traffic complexity by using the geometric-based intrinsic complexity measure called König metric. The demonstration of König metric shows the potential that the algorithm can capture the disorganized traffic which represents the difficulty of maintaining situational awareness as expected by the air traffic controller. We reformulate the optimization problem with two trajectory separation approaches including delaying flight departure time and allocating the new flight level subject to limited delay time of departure, limited changes of flight levels and fuel consumption constraints. We propose our solution to solve daily traffic demands in the regional French airspace. The resolution process uses the distributed metaheuristic algorithm to optimize aircraft trajectories in 4D environment with the objective of finding the optimal air traffic complexity. The experimental results shows the reduction of maximum complexity more than 95% with average delay of 2.69 minutes. The optimized trajectories can save fuel more than 80000 kg. The proposed algorithm not only reduces the air traffic complexity but also maintain its distribution in traffic. The research results represent further steps towards taking other trajectory separations methods and aircraft trajectory uncertainties into account, developing our approach at the continental scale as well as adapting it in the pre-tactical and tactical planning phase. Civil Aviation Authority of Singapore (CAAS) Accepted version This research is partially supported by NTU-CAAS Research Grant M4062429.052 by Air Traffic Management Research Institute, School of Mechanical and Aerospace Engineering, Nanyang Technological University, Singapore.

Drone has become more and more popular in various civil applications due to the open of the low-altitude airspace and its easy operation. Unlike the common search and track task for the target, the new search and lock (SAL) task is focused on in this paper. In the SAL task, multiple drones first try to detect the moving ground target cooperatively. Then they must lock the target by covering all the surrounding area of it at a low flight altitude, which indicates that the target can be watched clearly in all directions from then on. The SAL task can be applied in the panorama shot for the moving ground target. First, the low-altitude urban airspace is discretized into cubes, based on which the flight rules of drone are defined. The field of view (FOV) of drone is modeled considering the flight altitude and the block of buildings. For the cooperation among multiple drones, the constraints on the type of waypoint, the communication distance and the collision avoidance are all included. The goal in the search phase is to cover more area which have not been visited recently to increase the probability of detecting the target, and it is expected to lock the target as soon as possible in the lock phase. A new swarm-based imitative learning optimization (SBILO) algorithm is proposed to determine the waypoint of drone in the search phase considering the characteristic of the established SAL model. To have a quick response to the escape behavior of the target in the lock phase, the waypoint of drone is generated in a distributed way to cover more surrounding area of the target and lock it gradually. The case of losing the target in the FOV of all drones is also addressed by covering more possible places where the target may appear. Simulation results demonstrate that the SAL task can be performed efficiently by the drones with the flight routes obtained by the proposed SBILO algorithm and the distributed asynchronous decision-make (DADM) approach. Ministry of Education (MOE) Submitted/Accepted version This research work is supported by the Chongqing Research Program of Basic Research and Frontier Technology with the grant numbers of cstc2020jcyjmsxmX0602, Fundamental Research Funds for the Central Universities with the project reference number of 2020CDJ-LHZZ-066, China Scholarship Council with the project reference number of 201906055030. This collaborative research is also supported by the Ministry of Education (MOE, Singapore) Tier-1 project research grant (Project ID: 2018-T1-002-124) and the UAS Program on “Urban Aerial Transport Traffic Management and Systems” in the ATMRI, NTU, Singapore.

The use of intelligent decision aids in Air Traffic Management is recommended to manage the exponential rise in global air traffic. Consequently, trust, which is one of the main drivers of how Air Traffic Controllers use such decision aids, has become an important area of research. It has been suggested that there is a strong emotional influence in the formation of Human-Human Trust, and it is unclear if this paradigm is valid for Human-Autonomy Trust. The extent of the cognitive and emotional components in the initial trust relationship between Air Traffic Controllers and a simulated conflict detection autonomous decision aid was examined with the use of functional Magnetic Resonance Imaging. The results confirmed that the emotional component showed higher activation than the cognitive component for the initial formation of trust between Air Traffic Controllers and autonomous decision aids. Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme, with Research Grant 04SBS000704C160. IRB approval for this study was granted by the Nanyang Technological University Institutional Review Board (IRB) (NTU IRBIRB-2018-12-002).

Detection of air traffic conflicts in a weather constrained airspace is challenging given the inherent uncertainties and aircraft maneuvers which give rise to new conflict birth-points. Traditional conflict detection tools are untenable in such situations as they primarily rely on flight-plan, aircraft performance characteristics and trajectories projection in short-term (2-4 minutes). This work adopts a convolutional neural network (CNN) model, on radar-like images, for conflict detection task in a constrained airspace. The CNN models are well-known for their learning capabilities when dealing with unstructured data like pixelated images. In this study, historical ADS-B data with weather constrained airspace is input as pixelated images to the CNN model. The learned model was compared with two well-known models for conflict detection (CD). The results demonstrated that the CNN based model was able to predict off-nominal conflict with high accuracy. The CNN model also demonstrated its ability to predict off-nominal conflict early for a given ten-minute look-ahead window. The CNN based model also showed low levels of false alarm signals as compared to other models. Generally speaking, all models showed low probabilities of miss-detection, mostly in the early phase of the 10-minute look-ahead window. This novel approach may serve to develop effective CD algorithms with longer look-ahead time and may aid in early detection of air traffic conflicts in non-nominal scenarios. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research was supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

In this paper, a constrained image-based visual servoing control method for the shipboard landing problem of unmanned helicopters is proposed. First, the pitch and roll motion of ship are predicted by an auto-regressive (AR) model to determine an appropriate period for landing. Subsequently, a novel robust sliding mode controller without linear velocity measurements is developed on the basis of the perspective image feature in a virtual image plane. Meanwhile, a modified Chebyshev Neural Network (CNN) is proposed to estimate the uncertainties including the linear acceleration of ship and translational perturbation, while an adaptive law is employed to compensate the influence of rotational disturbances. The whole controller only requires the measurements feedback of a vision sensor and an inertial measurement unit (IMU). Ulteriorly, to prevent the visual target on the ship from going beyond the field of view of camera, the constrained controller is developed by a control barrier function and a quadratic programming, where the unknown relative velocity is estimated by a velocity observer. Finally, simulations are implemented to substantiate the capability of the presented shipboard landing control method. Accepted version This work was supported by the Beijing Natural Science Foundation (no. 4202038), China Postdoctoral Science Foundation (no. 2020TQ0028), the National Natural Science Foundation of China (no. 61827901), the National Key R&D Program of China (no. 2018YFC1506401), China Scholarship Council, and NTU ATMRI Project Agreement Research Grant (no. 2016-01).

Aircraft trajectory prediction is a challenging problem in air traffic control, especially for conflict detection. Traditional trajectory predictors require a variety of inputs such as flight-plans, aircraft performance models, meteorological forecasts, etc. Many of these data are subjected to environmental uncertainties. Further, limited information about such inputs, especially the lack of aircraft tactical intent, makes trajectory prediction a challenging task. In this work, we propose a deep learning model that performs trajectory prediction by modeling and incorporating aircraft tactical intent. The proposed model adopts the encoder-decoder architecture and makes use of the convolutional layer as well as Gated Recurrent Units (GRUs). The proposed model does not require explicit information about aircraft performance and wind data. Results demonstrate that the provision of enriched aircraft intent, together with appropriate model design, could improve the prediction error up to 30% at a prediction horizon of 10 minutes (from 4.9 nautical miles to 3.4 nautical miles). The model also guarantees the mean error growth rate with increasing look-ahead time to be lower than 0.2 nautical miles per minute. In addition, the model offers a very low variance in the prediction, which satisfies the variance-standard specified by EUROCONTROL (EU Organization for Safety and Navigation of Air Traffic) for trajectory predictors. The proposed model also outperforms the state-of-the-art trajectory prediction model, where the Root Mean Square Error (RMSE) is reduced from 0.0203 to 0.0018 for latitude prediction, and from 0.0482 to 0.0021 for longitude prediction in a single prediction step of 15 seconds look-ahead. We showed that the pre-trained model on ADS-B data maintains its high performance, in terms of cross-track and along-track errors, when being validated in the Bluesky Air Traffic Simulator. The proposed model would significantly improve the performance of conflict detection systems where such trajectory prediction models are needed. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Published version This work was supported in part by the National Research Foundation, Singapore; and in part by the Civil Aviation Authority of Singapore under the Aviation Transformation Program.

Undetected partial actuator faults on multi-rotor UAVs can lead to system failures and uncontrolled crashes, necessitating the development of accurate and efficient fault detection and isolation (FDI) strategy. This paper proposes a hybrid FDI model for a quadrotor UAV that integrates an extreme learning neuro-fuzzy algorithm with a model-based extended Kalman filter (EKF). Three FDI models using Fuzzy-ELM, R-EL-ANFIS, and EL-ANFIS are compared based on training, validation performances, and sensitivity to weaker and shorter actuator faults. They are also tested online for linear and nonlinear incipient faults by measuring their isolation time delays and accuracies. The results show that the Fuzzy-ELM FDI model exhibits greater efficiency and sensitivity, while Fuzzy-ELM and R-EL-ANFIS FDI models demonstrate better performance than a conventional neuro-fuzzy algorithm, ANFIS. Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of the Civil Aviation Authority of Singapore. The PhD candidature scholarship provided to the first author by Nanyang Technological University through Air Traffic Management Research Institute Leader’s Track is greatly appreciated.

The frequent intrusion of unmanned aerial vehicles (UAVs) into civil aviation airports poses a severe challenge to the safety management of the airport. The subsequent flight delays and even the closure of the runway have seriously affected the normal operation of the airport. To measure the collision risk of the Unmanned-Manned (U-M) encounter (intruding UAV to manned aircraft), a collision risk assessment platform is developed in this paper to predict the U-M encounter collision risk of the airport restricted area. The designed platform can be used to dynamically generate 3D trajectories of the U-M encounter, and to assess the collision risk with the changing of the relative distance in the probabilistic curve. Further, to predict the collision risk of the airport restricted area, the collision risk heatmap representation is then introduced. A case study of Changi Airport, Singapore is employed to generate the runway collision risk heatmap based on the designed collision risk assessment platform. The results demonstrate how the collision risk of the separate runway and the integrated runway distributes. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This project is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

With the broader adoption of digital air traffic control towers, real-time video data is expected to complement the current surveillance system (if available) and improve airport performance in terms of safety and efficiency. However, to fully utilize such data, a suite of computer vision algorithms needs to be developed for extracting useful information from real-time video feeds. Currently, most of the studies in the literature have focused only on the detection and tracking of aircraft on the airport surface, while approaching aircraft also play an essential role in airport and runway operations. The distance-to-touchdown of approaching aircraft is a critical parameter in final approach spacing and departure sequencing. Therefore, this research proposes a deep learning approach for estimating the distance of approaching aircraft to touchdown using multi-view video feeds. The proposed approach adopts a state-of-the-art computer vision model with an auto-calibration technique for detecting the approaching aircraft and extracting feature vectors from multiple camera views under various lighting and weather conditions. Then, an ensemble approach is introduced for combining the input vectors for distance estimation. The approach is evaluated with both Changi Airport simulated and real video data. Firstly, the proposed approach is designed to be easily updated and adapted for different camera system configurations. Secondly, the proposed approach has successfully combined the strength of both monoscopic and stereoscopic approaches to provide accurate distance-to-touchdown prediction in various scenarios. The experimental results demonstrate the advantages of the proposed approach with stable performance and low predicted errors (Mean Absolute Percentage Error = 0.18%) in estimating the distance-to-touchdown up to 10 NM. Such capability in a Digital Tower environment can augment the runway controller’s sequencing and final approach spacing capabilities. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Published version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

Tracking systems are a core component of future UAS Traffic Management (UTM) systems. This study aims at providing technical fundamentals for supporting the robust standardization of UTM tracking systems. To evaluate the impact of tracking system performances on the effectiveness of centralized conflict detection, we propose a simulation-based quantitative method. Monte Carlo simulations were used to acquire data for tracking systems behaviors under various operational uncertainties for a pair-wise flight encounter scenario. Sensitivity analysis of safety performance in managing Well-Clear Violations (WCVs) were performed with respect to performance parameters (transmission availability, UAS-side and ground side position update rates and transmission latency) for a cooperative tracking system. It was found that extrapolation techniques help to mitigate long intervals between successive UA position updates. UAS position and velocity update rates over 1 Hz are recommended. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not reflect the views of National Research Foundation, Singapore and the Civil Aviation Authority of Singapore.

The air traffic control paradigm is shifting from sector-based operations to cross-border flow-centric approaches to overcome sectors' geographical limits. Under this paradigm, effective air traffic flow coordination at flow intersections is crucial for efficiently utilizing available airspace resources and avoiding inefficiencies caused by high demand. This paper proposes a dynamic air traffic flow coordination framework to identify, predict, assess, and coordinate the evolving air traffic flows to enable more efficient flow configuration. Firstly, nominal flow intersections (NFI) are identified through hierarchical clustering of flight trajectory intersections and graph analytics of daily traffic flow patterns. Secondly, spatial-temporal flow features are represented as sequences of flights transiting through the NFIs over time. These features are used to predict the traffic demand at the NFIs during a given future period through a transformer-based neural network. Thirdly, for each NFI, the acceptable flow limit is determined by identifying the phase transition of the normalized flight transition duration from its neighboring NFIs versus the traffic demand. Finally, when the predicted demand exceeds the flow limit, by evaluating the available capacity at different NFIs in the airspace, the flow excess is alternated onto other NFIs to optimize and re-configure the air traffic demand to avoid traffic overload. An experimental study was carried out in French airspace using the proposed framework base on the ADS-B data in December 2019. Results showed that the proposed prediction model approximated the actual flow values with the coefficient of determination ($R^2$) above 0.9 and mean absolute percentage error (MAPE) below 20%. Acceptable flow limit determination showed that for above 68% NFIs, the flight transition duration increases sharply when the demand exceeds a certain level. The flow excess at an NFI whose demand was predicted to exceed its limit was coordinated, and the potential increase in the flight transition duration caused by the flow excess was avoided. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Founda- tion, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme

Wake hazard is one of the factors affecting the separation requirement for same-track fixed-wing aircraft in civil aviation. It could also be a contributing factor for UAS traffic management (UTM) with multirotors, especially with the widespread use of ADS-B like remoteID for UAS that greatly reduces the separation requirements due to communication, navigation, and surveillance (CNS) performance of the aircraft and the airspace. While the wake vortex structure and decay mechanism for fixed-wing aircraft has been well studied over the years, the same dynamics might not be observed in multirotor aircraft due to difference in lift generation mechanism. Additionally, the difference in wake vortex generation from different multirotor configurations were not as well studied. This paper presents the simulation results from two quadrotors (DJI Phantom 3 size, ~ 1.2 kg and Inspire 1 size, ~ 3.5 kg) and one hexarotors (DJI M600 sized, ~ 15 kg) for a fixed follow distance behind those aircraft. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This project is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme on Unmanned Aircraft Systems (UAS) in the topic of Separation Minima. The support from Air Traffic Management Research Institute (ATMRI) and the School of Mechanical and Aerospace Engineering (MAE), Nanyang Technological University (NTU) is also appreciated. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not reflect the views of National Research Foundation, Singapore and the Civil Aviation Authority of Singapore. The OpenFOAM based simulations were made possible by the computational resource allocated by the National Supercomputing Centre (NSCC) under Project account 12001778 and 12002591.

With the promise to revolutionize many industries, the popularity of Unmanned Aerial Vehicles (UAVs) has surged in the last two decades. The use of multirotor UAVs in urban areas is gaining traction, with applications such as cargo delivery, emergency response, and recreational flight. However, unlike manned aviation, UAVs do not have a standardized design, testing, and certification procedures. Hence, a lot of countries raised their concerns about the risk of potential multi-rotor UAV crashes and impacting pedestrians on the ground. Apart from high-rise buildings, covered linkways are a ubiquitous sight in countries like Singapore with a primary role to provide shelter for pedestrians in the event of rain. Therefore, this study investigates the collision severity of a multirotor UAV impacting a covered linkway. Briefly put, the level of protection that covered linkways can provide to pedestrians from a multirotor UAV falling from height is analyzed. The methodology in this study is divided into three phases. In Phase 1, Finite Element Analysis (FEA) is employed to simulate the impact of two specific multirotor UAV designs and solid spheres with comparable material properties as they collide onto a simplified covered linkway panel using ABAQUS/CAE. Subsequently, parametric analysis using Python scripting is performed to assess the collision severity due to variations in mass and crash speed of rigid sphere. Phases 2 and 3 explore the impact of the honeycomb structure modelling on the impact simulation. Findings from this study will be beneficial for the computation of sheltering factor in UAV ground risk assessment. Submitted/Accepted version

Small unmanned aircraft systems or drones are expected to be used for different applications, such as parcel delivery, inspection, and aerial photography, in urban areas. However, drones usually use an electric system to power up the propulsion, communications, navigation, and flight control system, which means that it is not as reliable as the manned aircraft system and may result in failure during operation and then crash to the ground. At present, there is almost no extensive publication about the high-fidelity modeling used by drones to calculate the crash trajectory and point of crash. The experimental data for modeling and simulation verification of multirotor aircraft are limited. So far, crash trajectory prediction has been limited to point mass or ballistic methods, and these methods are usually only suitable for complete power failure and without any control system. This study intends to investigate the effects of different multirotor drones’ failure modes on its crash trajectory and crash area compared to the ballistic model by using ADAMS and MATLAB co-simulation methods. Conclusions from the study show the crash trajectory, flight distance, and impact speed of the drones under four failure modes, which are quite different from the ballistic trajectory. The findings can potentially contribute to better risk assessment of the multirotor drones for the urban environment operation. Submitted/Accepted version

With the increasing interest in the utilization of urban airspace for unmanned aerial system (UAS) operations, be it for cargo delivery or passenger transportation, the future city sky could become quite a crowded place if all those visions become reality. In such case, the question of how close two operations could be conducted while sharing an airspace must be answered to ensure the safe and efficient use of the urban airspace. The lateral separation needed to prevent inadvertent intrusion by neighboring tracks is especially important when designing airspace corridors constrained by the urban landscape. This paper documents a series of flight tests conducted in open fields with the goal of assessing the along-track (longitudinal), cross-track (lateral),and altitude (height) deviation under two different flight conditions: operator guided operation within Visual Line of Sight (VLoS), and the waypoint-guided mission analogous to operating Beyond Visual Line of Sight (BVLoS). The flight test statistics were also compared to the Monte-Carlo based path prediction model that was used in earlier studies for collision prediction. The goal is to determine the Flight Technical Error (FTE) that forms a part of the Total System Error (TSE) for Performance Based Navigation (PBN) of UAS to support the establishment of separation requirements in urban airspace. Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University Accepted version This project is supported by the Civil Aviation Authority of Singapore and the Nanyang Technological University, Singapore under their collaboration in the Air Traffic Management Research Institute.

With increasing air traffic, there is an ever-growing need for Air Traffic Controllers (ATCO) to efficiently manage traffic and congestion. Congestion often leads to increased delays in the Terminal Maneuvering Area (TMA), causing large amounts of fuel burn and detrimental environmental impacts. Approaches such as the Extended Arrival Manager (E-AMAN) propose solutions to absorb such delays, whereby flights are scheduled much before they enter the TMA. However, such an approach requires a speed management system where flights can coordinate to absorb system-level delays in their en-route phase. This paper proposes a Multi-Agent System (MAS) approach using Deep Reinforcement Learning to model and train flights as agents which can coordinate with each other to effectively absorb system-level delays. The simulations utilize Multi-Agent POsthumous Credit Assignment in Unity and test two reward approaches. Initial findings reveal an average of 3.3 minutes of system-level delay absorptions from a required delay of 4 minutes. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This project is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

During peak-hour busy airports have the challenge of turning aircraft around as quickly as possible, which includes sequencing their landings and take-offs with maximum efficiency, without sacrificing safety. This problem, termed aircraft sequencing problem (ASP) has traditionally been hard to solve optimally in real-time, even for flights over a one-hour planning window. In this article, we present a novel data-splitting algorithm to solve the ASP on a single runway with the objective to minimize the total delay in the system both under segregated and mixed mode of operation. The problem is formulated as a 0–1 mixed integer program, taking into account several realistic constraints, including safety separation standards, wide time-windows, and constrained position shifting. Following divide-and-conquer paradigm, the algorithm divides the given set of flights into several disjoint subsets, each of which is optimized using 0–1 MIP while ensuring the optimality of the entire set. One hour peak-traffic instances of this problem, which is NP-hard in general, are computationally difficult to solve with direct application of the commercial solver, as well as existing state-of-the-art dynamic programming method. Using our data-splitting algorithm, various randomly generated instances of the problem can be solved optimally in near real-time, with time savings of over 90%. Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University This research has been partially supported under ATMRI (NTU-CAAS) Grant No. M4061216.

Drones should be allowed to respond to dynamic urban environments and self-adjust their paths, safely and efficiently. Existing studies fail to develop a comprehensive approach to deal with drone encountering various dynamic changes over the course of flying. In this paper, an adaptive path replanning (APReP) method is proposed in discrete urban environments by considering the features of different types of dynamic changes, and the coordination among drones as well. First, various dynamic changes are concluded into three types. Three strategies are developed to conduct the path replanning for a single-drone operation under different combinations of dynamic changes. As the path replanning is extended to the operation involving multiple drones, the orders of planning are determined by task priority, path planning strategy and competition mechanism. A discrete rapidly-exploring random tree (DRRT) algorithm is presented to generate the path considering the characteristic of discrete urban environments. Simulation results demonstrate that DRRT algorithm is suitable for the path replanning problems considered, and the three proposed path replanning strategies are valid to cope with the corresponding types of dynamic change. Compare to other two algorithms, APReP algorithm is more efficient in large-scale problems with a number of dynamic changes. Accepted version This work was supported in part by the Fundamental Research Funds for the Central Universities under Project Reference 2019CDJGFHK001, in part by the China Scholarship Council under Project/reference 201906055030, and in part by the UAS Programme’s research grants, ATMRI, NTU, Singapore.

In mixed-mode operation, arrivals and departures are allowed to land and depart on the same runway. An appropriate strategy from air traffic controllers for arrivals and departures sequencing would boost the runway throughput significantly. On the other hand, safety is still the most crucial feature in the operation. Therefore, to assist air traffic controllers to make decisions on departures and arrivals with efficient utilization of runway capacity and safe operations, this paper proposed a Multi-agent Deep Reinforcement Learning approach using Multi-agent Deep Deterministic Policy Gradient to train two agents simultaneously: departure agent, and arrival agent. The departure agent makes departure slotting decisions for departures while the arrival agent determines the time delay or spacing decision on the arrival stream. A data-driven simulation environment is developed using Singapore Changi Airport data to support the learning process. Besides, a random sampling technique is also introduced to reduce redundant samples and increase off-policy sample efficiency. Moreover, the impact of different reward functions on runway throughput is also investigated and two specific models, e.g., 'arrival priority' and 'departure priority', are selected for further analysis in this study. As the result, by comparing the trained models with the ad-hoc model, the proposed approach increases the runway throughput significantly, with the highest 12.8% additional departure releases and overall 5.3% additional departure releases in identical environments while safety separations are maintained. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

The Unmanned Aerial Vehicles (UAVs) have gained increasing attention recently, and depth estimation is one of the essential tasks for the safe operation of UAVs, especially for drones at low altitudes. Considering the limitations of UAVs’ size and payload, innovative methods combined with deep learning techniques have taken the place of traditional sensors to become the mainstream for predicting per-pixel depth information. Since supervised depth estimation methods require a massive amount of depth ground truth as the supervisory signal. This article proposes an unsupervised framework to tackle the issue of predicting the depth map given a sequence of monocular images. Our model can solve the problem of scale ambiguity by training the depth subnetwork jointly with the pose subnetwork. Moreover, we introduce a modified loss function that utilizes a weighted photometric loss combined with the edge-aware smoothness loss to optimize the training. The evaluation results are compared with the model without weighted loss and other unsupervised monocular depth estimation models (Monodepth and Monodepth2). Our model shows better performance than the others, indicating potential assistance in enhancing the capability of UAVs to estimate distance with the surrounding environment. Civil Aviation Authority of Singapore (CAAS) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

The wind is one of the major factors that may cause unmanned aerial vehicles (UAVs) to crash and pose fatality risk to the population and property damage risk to infrastructures. This paper investigates wind patterns on temporal scales to identify high-risk periods in terms of wind conditions for safe UAV operations in urban airspace. The research starts with the historical wind speed data analysis using statistical approaches. As the wind speed data does not follow normal distribution after checking, a nonparametric approach of the Kruskal-Wallis test is applied for hypothesis testing to see if there is a significant difference in the median wind speed in different years. Regression analyses are also performed for monthly wind speed data to check any significant trends that could facilitate the predictions of average wind speed in the long term. This study will contribute to safe air traffic management for UAV operations in low-altitude urban airspace by mitigating adverse wind effects. Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme. The Research Student Scholarship (RSS) provided by the Nanyang Technological University (NTU) to the second author is also acknowledged.

The activity of generic flight path monitoring of an aircraft on the radar screen is one the major problem which results in air traffic controllers (ATCOs) losing awareness of it. A novel eye signature, known as the radar command group (RCG) time entropy signature, is thus presented in this paper. This signature seeks to model the monitoring behavior of ATCOs using eye-tracking technique during this activity. Acquisition and representation of this monitoring behavior is achieved by first identifying the respective fixation count and mouseover label movement on the aircraft to establish a RCG. The regularity of a RCG exhibited on an aircraft is then calculated using the information entropy formula. Real time simulations are conducted for 88 one-hour experimental sessions with licensed and non-licensed participants comprising of three expertise levels, using scenarios that mimic actual air traffic, consisting of various flight path configuration. Test results of the RCG time entropy signature showed that, an aircraft flying on a flight path with longer distance and no change in altitude likely results in a more regular generic flight path monitoring. Furthermore, it can also be used to differentiate the novice participants from intermediates and experts, as novices lack the training. An ATCO's expertise level can thus be determined and benchmarked accordingly, allowing this signature to be applied on future ATCO training. Economic Development Board (EDB) Accepted version This work was supported in part by Thales Solutions Asia Pte Ltd, under the Economic Development Board, Industrial Postgraduate Programme, with Nanyang Technological University, Singapore.

This work aims to enhance the safety of air traffic in a procedural airspace by air traffic flow management (ATFM) without compromizing air traffic demand. In civil aviation, collision risk is an important indicator for air traffic safety assessment. In this work, we propose a generic ATFM framework based on multiobjective optimization for reducing lateral collision risk of the traffic in a given procedural airspace. The proposed framework aims to optimize the flight level assignment for a given set of flight plans such that the lateral collision risk can be reduced. To achieve this goal, we formulate an optimization problem containing two partially conflicting objective functions. We then adopt three well-known evolutionary algorithms, i.e., MODPSO, NSGA-II, and MOEA/D to solve the optimization problem. We specially design some of the operators of those al- gorithms to make them suitable for the optimization problem. We merge the solutions yielded by those three algorithms and filter out the final Pareto solutions. We carry out a case study on the procedural airspace of Singapore flight information region (FIR) with respect to twelve daily traffic data selected from the real traffic data for December 2019. Experiment results demonstrates that the lateral occupancy which is the key contributor to lateral risk can be reduced by 10.65% to 93.05% at a strategic planning level. This research contribute to strategic flight planning by assigning flight levels that may reduce the risk of collision in procedural airspace. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

Unmanned aircraft systems (UAS) are anticipated to increase in the future and be employed for a variety of applications such as parcel delivery, structural inspections, aerial photography, and surveillance. However, the growth of the UAS industry is impeded by delayed operational approvals and risk assessments. This is likely due to the limited knowledge on the third-party risk (TPR) of UAS operation in the environment. For metropolitan environments like Singapore, the operation of UAS pose high risk to the population on ground due to high population density. Hence, it is of vital importance to assess the ground risk of UAS operations and to subsequently demarcate low, medium, and high areas within the city. In our preliminary study, a ground risk framework, adapted from Joint Authorities for Rulemaking of Unmanned Systems (JARUS) Specific Operations Risk Assessment (SORA), and a semi-quantitative ground risk matrix was established to quantitatively generate UAS failure rate thresholds as well as perform ground risk-tiering of neighborhoods. Ground risk mapping was illustrated for UAS of four different maximum take-off weight (MTOW). It was observed that larger UAS had increased ground risk especially in areas with highly dense population. Study of ground risk in urbanized environments will provide insights on the level of safety required for UAS operations. This could aid aviation authorities in formulating comprehensive risk assessments and reduce the time required to approve UAS operations in metropolitan environments. However, it is noteworthy that further studies are required to validate the results of the risk matrix. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

With the rising adoption of multi-rotor UAVs, it has become ever more crucial that their airworthiness is ensured, especially for hobby-grade UAVs. For some UAVs, their onboard components might not be as reliable as required and as such, faults can occur during flight operations and may develop to cause catastrophe. Hence, these faults need to be accurately diagnosed and quickly mitigated. This paper presents a fault diagnosis model for a quadrotor subjected to partial actuator faults. Flight simulations, with actuator and GPS sensor fault injections, are performed on a hardware-in-the-loop experimental setup to gather flight data consisting of multiple sensors. Based on this data, a preliminary controllability threshold analysis is conducted for the quadrotor. After that, a fault diagnosis model using an online sequential fuzzy-extreme learning machine (OS-Fuzzy-ELM) is trained to locate the actuator faults on the quadrotor UAV. The trained model presents an average testing accuracy and macro-averaged F1 score of 80.2% and 78.0%. A subsequent study to isolate sensor and actuator faults presents the testing accuracy and macro-averaged F1 score to be 1.95% and 1.21%, marginally better than a fault diagnosis model based on a single-layer feedforward network Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation (NRF), Singapore, and the Civil Aviation Authority of Singapore (CAAS), under the Aviation Transformation Programme (ATP) on “Integration of Unmanned Aircraft Systems (UAS) into the Airspace." Any opinions, findings, and conclusions, or recommendations expressed in this material are those of the authors and do not reflect the views of the Civil Aviation Authority of Singapore. Also, the Ph.D. candidature scholarship provided to the first author by Nanyang Technological University (NTU, Singapore) through Air Traffic Management Research Institute (ATMRI) Leader’s Track is greatly appreciated

Urban air mobility (UAM) has been attracting stakeholders' attention for its great potential in playing an important role in future urban traffic systems. Expected high density of UAM operation requires advanced management strategy of airspace resources. In this paper, we present an Airspace-Resources-Centric (ARC) approach for UAM flight planning. In the first place, an ARC concept for urban airspace management is developed. Then a method for 4D trajectory planning with optimal airspace utilization was designed under the concept. The method consists three steps. Firstly, a flight level is assigned to each flight plan. Secondly, an optimal 3D path is generated by A* algorithm and is smoothed. Lastly, 4D trajectory generation is modeled as a mathematical optimization problem, and is solved by a commercial solver. Numerical simulation results show that the method plans trajectory for UAM with less than 1 second computational time per flight. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not reflect the views of National Research Foundation, Singapore and the Civil Aviation Authority of Singapore.

Air route crossing waypoint optimization is one of the effective ways to improve airspace utilization, capacity and resilience in dealing with air traffic congestion and delay. However, research is lacking on the optimization of multiple Crossing Waypoints (CWPs) in the fragmented airspace separated by Prohibited, Restricted and Dangerous areas (PRDs). To tackle this issue, this paper proposes an Artificial Potential Field (APF) model considering attractive forces produced by the optimal routes and repulsive forces generated by obstacles. An optimization framework based on the APF model is proposed to optimize the different airspace topologies varying the number of CWPs, air route segments, and PRDs. Based on the framework, an adaptive method is developed to dynamically control the optimization process in minimizing the total air route cost. The proposed model is applied to busy controlled airspace. And the obtained results show that after optimization the safety-related indicators: conflict number and controller workload reduced by 7.75% and 6.51% respectively. As for the cost-effectiveness indicators: total route length, total air route cost and non-linear coefficient, declined by 1.74%, 3.13% and 1.70% respectively. While the predictability indicator, total flight delay, saw a notable reduction by 7.96%. The proposed framework and methodology can also provide an insight in the understanding of the optimization to other network systems. Civil Aviation Authority of Singapore (CAAS) Published version This research was supported by the Civil Aviation Authority of Singapore and the Nanyang Technological University, Singapore under their collaboration in the Air Traffic Management Research Institute. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not reflect the views of the Civil Aviation Authority of Singapore.

Intractable delays occur in air traffic due to the imbalance between ever-increasing air traffic demand and limited airspace capacity. As air traffic is associated with complex air transport systems, delays can be magnified and propagated throughout these systems, resulting in the emergent behavior known as delay propagation. An understanding of delay propagation dynamics is pertinent to modern air traffic management. In this work, we present a complex network perspective of delay propagation dynamics. Specifically, we model air traffic scenarios using spatial-temporal networks with airports as the nodes. To establish the dynamic edges between the nodes, we develop a delay propagation method and apply it to a given set of air traffic schedules. Based on the constructed spatial-temporal networks, we suggest three metrics—magnitude, severity, and speed—to gauge delay propagation dynamics. To validate the effectiveness of the proposed method, we carry out case studies on domestic flights in the Southeastern Asia Region (SAR) and the United States. Experiments demonstrate that the propagation magnitude in terms of the number of flights affected by delay propagation and the amount of propagated delays for the US traffic are respectively five and 10 times those of the SAR. Experiments further reveal that the propagation speed for U.S. traffic is eight times faster than that of the SAR. The delay propagation dynamics reveal that about six hub airports in the SAR have significant propagated delays, while the situation in the United States is considerably worse, with a corresponding number of around 16. This work provides a potent tool for tracing the evolution of air traffic delays. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Accepted version This work was supported by SUG Research Grant M4082126.050 by School of Mechanical and Aerospace Engineering (MAE), Nanyang Technological University (NTU), Singapore and NTU-CAAS Research Grant M4062429.052 by ATM Research Institute, School of MAE, NTU, Singapore.

Small unmanned aerial vehicles (sUAV) or 'drones' are projected to rapidly develop in the future and be employed for a multitude of applications such as parcel delivery, search and rescue, surveillance monitoring, and farming. Consequently, an increase in sUAV traffic may lead to a higher likelihood of failure, and crash occur primarily due to environmental factors. Hence, the objective of this study was to investigate the influence of environmental factors on sUAV operations.The study identifies three environmental risk categories relevant to Singapore’s context, namely weather, obstacle, and signal. Subsequently, Python code was developed to perform data analysis on environmental risk data gathered from open source, relevant agencies, and literature surveys. Three environmental risk factors were considered for proof of concept, namely rainfall, windspeed, bird population, and cell tower signal strength.From the data analysis, the average daily rainfall across weather stations in Singapore was observed to be 8.58 mm. In August, weather stations in West of Singapore, such as Choa Chu Kang and Tengah, reported the highest rainfall, approximately 22.65 mm. In contrast, weather stations in the East such as Changi and Pasir Ris reported the lowest rainfall near 0.036 mm in February. Data of the three most invasive bird species in Singapore was examined. The site of Chinese Garden in West Singapore reported the highest bird population of 432 birds per site. The study of bird population data potentially provides an insight into the possibility of sUAV bird strike events. Similarly, signal strength data from cell towers in Singapore was studied.Overall environmental risk-tiering can be visualized using all the analyzed environmental factors data on a risk map. The regions with high rainfall, increased bird population per site and strong signal interference can be extracted.These high-risk locations can be expected to be more hazardous for sUAVs to fly in. Findings for environmental risk-tiering and risk-mapping are preliminary and based on the available environmental data as identified in this study. Further research will be done to include other environmental risk factors and to generate better accuracy of the environmental risk-tier map. Insights from this study will facilitate the demarcation of low, medium, and high-risk areas for risk assessment of sUAVs operation in Singapore. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

UAS navigational error is one of the concerns leading to non-conformance of assigned flight volume. This is especially noticeable in urban environment where global navigation satellite system (GNSS) positioning capability are affected by influencing factors such as signal blockage and multipath errors caused by surrounding objects like buildings, trees, and other obstructions [1]. This paper presents a study on how the vertical navigation performance is affected by the cause of change in the sky openness ratio (i.e., GNSS signal blockage) above the drone when a self-built UAS is operating near several types of urban obstructions in flight tests. Test sites are carefully selected to ensure they are representative to urban environments, which includes 1) open space carpark with a few short trees, 2) running parallel to row of buildings, 3) surrounded by buildings (narrow urban canyon), and 4) constraint area with a building and some tall trees. The data will be analysed to obtain the vertical position error with considerations for (i) Number of visible satellites versus flight altitude, (ii) Sensor-reported vertical co-variance (EPV), (iii) Vertical position errors at various locations by comparing the reported position (GNSS) to a reference value, and (iv) Factors contributing to vertical positional errors at various locations, (v) The influence of sky openness to the vertical position error. The measurement of sky-openness ratio is to determine whether 3D city model of the flight area is sufficient to predict the vertical navigation performance that could be expected. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

Air traffic is operated in an air traffic network (ATN) environment. It is pertinent to improve the robustness of ATNs as they are frequently exposed to manifold uncertainties which can break down their functioning components. Existing studies improve the robustness of an ATN by either rewiring its links or adding more ones. In this paper we discover the robustness paradox phenomenon in ATNs. Specifically, we claim to improve the robustness of an ATN by removing its links. In order to determine the links whose removal can improve an ATN's robustness, we develop a bi-objective optimization model with one objective maximizing the network's robustness and the other one minimizing the number of links to be removed. We further apply and modify a non-dominated sorting genetic algorithm (NSGA-II) to optimize the developed model. We then carry out experiments on nine real-world ATNs to validate the effectiveness of the proposed idea. We also compare the modified NSGA-II algorithm against NSGA-III, and MODPSO, which are famous and efficient multiobjective evolutionary algorithms. Experiments indicate that NSGA-II outperforms the compared algorithms and that robustness paradox phenomenon does exist in ATNs. This work provides a new perspective for aviation decision makers to better design and manage ATNs. Accepted version

It is an effective way to execute a complicated mission by cooperating unmanned vehicles. This paper focuses on a search- and-track (SAT) mission for an underwater target, and the mission is implemented by combining an unmanned aerial vehicle (UAV), an unmanned surface vehicle (USV) and an autonomous underwater vehicle (AUV). In the cooperative path planning model, the mission is divided into the search phase and the track phase, and the goals of the two phases are to maximize the search space and minimize the terminal error respectively. The constraints contain the maneuverability of vehicles and communication ranges between vehicles. Strategies based on random simulation experiments and asynchronous planning are developed to design the cooperative path planning algorithm in the two phases, and the paths are generated by an improved particle swarm optimization (IPSO) algorithm in a centralized or a distributed mode. Simulation results demonstrate that the proposed method can deal with different situations. The UAV&USV&AUV system is superior to the USV&AUV system in the SAT mission. Civil Aviation Authority of Singapore (CAAS) Accepted version This research work is supported by the Fundamental Research Funds for the Central Universities with the project reference number of 2019CDJGFHK001, China Scholarship Council with the project reference number of 201906055030 and the UAS Programme’s research grants, ATMRI, NTU, Singapore. The research attachment of the first author to the ATMRI provided by NTU’s School of Mechanical and Aerospace Engineering is also greatly appreciated.

Complex systems in the real world inevitably suffer from unpredictable perturbations, which can trigger system disasters, wreaking significant economical losses. To exploit the robustness of complex systems in the face of disturbances is of great significance. One of the most useful methods for system robustness analysis comes from the field of complex networks characterized by percolation theories. Many percolation theories, therefore, have been developed by researchers to investigate the robustness of diverse complex networks. Nevertheless, extant percolation theories are primarily devised for multilayer or interdependent networks. Little endeavor is dedicated to systems with multipartitivity structures, that is, multipartite networks, which are an indispensable part of complex networks. This paper fills this research gap by theoretically examining the robustness of multipartite networks under random or target node attacks. The generic percolation theory for robustness analysis of multipartite networks is accordingly put forward. To validate the correctness of the proposed percolation theory, we carry out simulations on computer-generated multipartite networks with Poisson degree distributions. The results yielded by the proposed theory coincide well with the simulations. Both theoretical and simulation results suggest that complex systems with multipartitivity structures could be more robust than those with multilayer structures. Accepted version

Air travelers' behavior is closely related to the operational performance of any airport terminal. Much of previous research has focused on how airport operators balance the number of facilities in a terminal and the Level of Service (LOS), while the behavior of passengers is less considered. Not much is known, however, about passenger's behavior during the entire departure process in an airport. In this study, we analyze empirical departure passenger's data to gain an insight into the regular patterns of their activities in an airport. We find that there exist two distinguished temporal patterns during two discretionary periods- post check-in and pre-security check, post security check and pre-boarding. The time that departure passengers spend in these two periods is well approximated by a double power-law distribution and an exponential truncated power-law distribution respectively. The two distinguished distributions suggest that there may be different mechanisms underlying passengers' behavior as indicated by previous studies on human mobility. We introduce a stochastic model that considers traveling experience and time pressure to capture the decision dynamics of human behavior. Simulation results suggest that traveling experience and time pressure dominate passenger's decisions before and after security respectively. Our findings contribute to a better understanding of human dynamics, and also offer the potential for optimizing and simulation of airport terminal operation. Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University Published version This work was supported in part by the National Natural Science Foundation of China under Grant U1833126, Grant 61773203, and Grant 61304190, and in part by the Civil Aviation Authority of Singapore and Nanyang Technological University, under their collaboration in the Air Traffic Management Research Institute.

As the use of multi-rotors is becoming common, understanding how the wake vortices develop and decay from a multi-rotor is important to reduce the effect of wake vortices on other multi-rotors and ensure their safety during operation by applying a wake separation minima. This paper proposes a method of flow simulation in ANSYS Fluent to study the decay of wake vortices from quadrotor propellers as the vortices evolve over distance and time. First, the near-field wake flow from the quadrotor propellers was simulated using Reynolds-Averaged Navier-Stokes (RANS) turbulence model. The velocity profile in the quadrotor frame of reference was then extracted to be applied for a flow simulation using Large Eddy Simulation (LES). The circulation strength as a function of distance could then be calculated, but it was limited by the length of the computational grid due to the wake flow being seen from the quadrotor frame of reference. To change into the stationary frame of reference, the interpolation data from the first stage LES was exported and modified by subtracting the freestream velocity from the velocity component in the freestream direction. The modified interpolation data was then used for the second stage LES, and the circulation strength in terms of time could be calculated from the resulting wake flow. Three cases of flow simulation are presented in this paper: first using 380-mm-diameter propellers based on the SUI Endurance quadrotor propellers with forward flight speed of 6 m/s, second using the 380-mm-diameter propellers with 12 m/s speed, and third using 240-mm-diameter propellers (similar to the propellers of DJI Phantom quadrotor) with 12 m/s speed. Findings regarding the decay of wake vortex from multi-rotor vehicles will help in determining the safe separation distance required to mitigate the hazards due to wake vortex encounters and ensure the safety of multi-rotor vehicles. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme on Unmanned Aircraft Systems (UAS) in the area of Separation Minima.

Drones or Unmanned Aerial Vehicles (UAVs) are expected to be used for different applications like parcel delivery, inspection, and aerial photography in the urban area. However, UAVs usually uses an electric system to power up the propulsion, communications, navigation, and flight control system, which means it is not as reliable as the manned aircraft system and may result in failure during operation and then crash to the ground. At present, there is almost no publication about the high-fidelity modeling used by UAVs to calculate the crash trajectory and point of crash. The experimental data for modeling and simulation verification of multi-rotor aircraft is limited. So far, crash trajectory prediction has been limited to point mass or ballistic methods, and these methods are usually only suitable for complete power failure and without any control system. This study intends to investigate the effects of different UAV failure modes on its crash trajectory and crash area compared to the ballistic model by using ADAMS and MATLAB co-simulation methods. Conclusions from the study show the crash trajectory, flight distance and impact speed of the UAV under four failure modes, which are quite different from the ballistic trajectory. The findings can potentially contribute to better risk assessment of the multi-rotor sUAV in the urban environment operation. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme on Unmanned Aircraft Systems (UAS) in the topic of Third-Party Risks. The support from the Air Traffic Management Research institute (ATMRI) and the School of Mechanical and Aerospace Engineering (MAE), Nanyang Technological University (NTU) is also appreciated.

With the increase interest in utilizing multirotorUAS in very lowaltitude airspace, the safety aspect associated with higher UAS density gains importance. One of the safety consideration is the separation requirements for UAS sharing the airspace, which consists of two major components: wake-based and CNS performance-based. To evaluate thewake-based separation requirements, thewake-field produced by the leadingUAS and the resulting encounter response by the follower UAS must be obtained. While previous studies on aircraft wake vortex evolution showed that the fuselage wake have small impact on the overall wake vortex strength and structures, the more distributed lift generation with less distance between the initial formation point of wake vortex to the fuselage could lead to a larger impact. This paper investigates the impact of including multirotor fuselage on simulating the strength and structure of the resulting wake vortex. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Accepted version This project is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme on Unmanned Aircraft Systems (UAS) in the topic of Separation Minima.

Airport departure metering can contain airside congestion but it may adversely impact scheduled gate assignments leading to passenger missed connections. Using an integrated landside airside framework, this study aims at evaluating the impact of departure metering on connecting flights and passenger connections. The proposed framework comprises of landside and airside simulators where the landside simulator simulates transfer passenger movements conditioned upon minimum connection time and the airside simulator simulates runway and taxiway movements based on planned operations and available data source. Departing aircraft movements are then metered to reduce taxi and take-off delays under uncertainties. Delayed arrivals and/or departures may lead to conflicts at gate rendering planned gate assignments infeasible or impractical. Using Singapore Changi airport A-SMGCS data, as a case study, it is found that DM may transfer as much as 12 minutes of waiting time from taxiways to gates. This leads to increased gate conflicts which additionally delays the gate-in time of an arriving aircraft by 2-7 minutes on average. Gate reassignments were found to significantly reduce both the number of gate conflicts and arrival queuing time at gates leading to less passenger delays. A minimum connection time of 70 minutes is found sufficient to reduce the probability of missed connections for transfer passengers. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of National Research Foundation, Singapore and the Civil Aviation Authority of Singapore.

The air traffic control paradigm is shifting from sector-based operations to cross-border flow-centric approaches to overcome sectors’ geographical limits. Under the flow-centric paradigm, prediction of the traffic flow at major flow intersections, defined as flow coordination points in this paper, may assist controllers in coordinating intersecting traffic flows which is the main challenge for implementing flow-centric concepts. This paper proposes to predict the flow at coordination points through a transformer neural network model. Firstly, the flow coordination points, i.e., the major flow intersections, are identified by hierarchical clustering of flight trajectory intersections whose location and connectivity characterize daily traffic flow patterns as a graph. The number of coordination points is optimized through graph analysis of the daily flow pattern evolution. Secondly, air traffic flow features in the airspace during a period are described as a “paragraph” whose “sentences” consist of the time and callsign sequences of flights transiting through the identified coordination points. Finally, a transformer neural network model is adopted to learn the sequential flow features and predict the future number of flights passing the coordination points. The proposed method is applied to French airspace based on one-month ADS-B data (from Dec 1, 2019, to Dec 31, 2019), including 158,856 flights. Results show that the proposed prediction model can approximate the actual flow values with a coefficient of determination (R2) between 0.909 to 0.99 and a mean absolute percentage error (MAPE) varying from 27.4% to 11.7% with respect to a 15-minute to 2-hour prediction window. The sustainability of the prediction accuracy under an increasing prediction window demonstrates the potential of the proposed model for longer-term flow prediction. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Published version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of National Research Foundation, Singapore and the Civil Aviation Authority of Singapore.

The increasing unmanned aircraft system (UAS) applications in urban environments pose challenges for safe and efficient low altitude air traffic management. As an essential enabler to meet these challenges, pre-flight 4D routes optimization is required to conduct conflict detection and resolution (CD&R) and to generate conflict-free flight routes before departure. Existing studies on strategic deconfliction cover several types of strategies such as scheduling or rerouting. However, a single type of strategy used to solve different types of conflicts may lead to an unsafe and inefficient way of conflict resolution. This paper proposes an adaptive decision-making framework to optimize the resolution strategies used for different types of conflicts with explainable mechanisms. The proposed framework is formulated as a double-layer optimization problem with the considerations of scheduling, speed adjustment, and rerouting strategies for conflict resolution. The first layer of the framework is established as a probabilistic selection model to make decisions on which strategy should be selected for what type of conflict. The second layer is developed as a mixed-integer nonlinear programming (MINLP) model to optimize the decision variables of the strategies selected by the first layer. To solve the proposed double-layer optimization problem, we introduce and improve a novel meta-heuristic stochastic fractal search (SFS) algorithm with two major improvements of a penalty-guided fitness function and an exploitation-exploration balancing scheme. Simulation results demonstrate that the proposed adaptive conflict resolution framework successfully optimizes the strategies used for each type of flight conflict, which subsequently optimizes the 4D routes with significant reductions in total operational cost, number of flight conflicts, and flight delays. The improved stochastic fractal search (ISFS) algorithm is also proved effective and reliable in solving the proposed optimization problem in different traffic density scenarios. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation (NRF), Singapore, and the Civil Aviation Authority of Singapore (CAAS), under the Aviation Transformation Programme (ATP).

Drone operations in low-altitude urban airspace might be influenced by weather conditions such as wind and rainfall. Severe weather conditions may exceed the threshold of drone tolerability and cause crash accidents, posing risks to people and property. To investigate the influence of weather conditions on drone operation, this paper presents a data-driven method for analysis of weather data and to identify different levels of risk for safe drone operations. Obtained weather data is first collected across Singapore’s 63 weather stations, and trend analysis are conducted to test if there are any significant trends in the yearly weather data. The risk standards in low-altitude urban environments are then classified based on the risk cost model. Similar levels of risk are clustered depending on the geographical location. Preliminary results show that the present weather data can be used to model our simulations as past historical weather data have no significant deviations. The results also concluded that low to high rainfall usually occurs at low wind speeds while high wind speeds tend to have low rainfall. The weather data analysis results can be used to generate an environmental risk-map for safe airspace planning and UAV path optimization. Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme. The Research Student Scholarship provided by the Nanyang Technological University to the second author is also acknowledged.

Various applications of unmanned aerial vehicles (UAVs) in urban environments facilitate our daily life and public services. However, UAV operations bring third party risk (TPR) issues, as UAV may crash to pedestrians and vehicles on the ground. It may also cause property damages to critical infrastructures and noise impacts to the public. Path planning is an effective method to mitigate these risks and impacts by avoiding high-risk areas before flight. However, most of the existing path planning methods focus on minimizing flight distance or energy cost, rarely considered risk cost. This paper develops a novel flight path optimization method that considers an integrated cost assessment model. The assessment model incorporates fatality risk, property damage risk, and noise impact, which is an extension of current TPR indicators at modeling and assessment levels. To solve the proposed integrated cost-based path optimization problem, a hybrid estimation of distribution algorithm (EDA) and CostA* (named as EDA-CostA*) algorithm is proposed, which provides both global and local heuristic in- formation for path searching in cost-based environments. A downtown area in Singapore is selected for the case study. Simulation results demonstrate the effectiveness of the developed cost-based path optimization model in reducing the risk cost. The statistical analysis for 100 sampled environments also shows the reliability of the proposed method, which reduced the cost by [42.64%, 44.15%] at 95% confidence level. Civil Aviation Authority of Singapore (CAAS) Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme on "Integration of Unmanned Aircraft Systems (UAS) into the Airspace." Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not reflect the views of National Research Foundation, Singapore and the Civil Aviation Authority of Singapore.

Conflicts between taxiing aircraft are resolved by making the aircraft with lower priority wait, slow down, or change their path. Prevalent priority assignment is based on rules such as First Come First Serve. However, this is not viable as a priority assignment done by an air-traffic controller (ATC) based on multiple factors. Thus, a machine learning approach is proposed to mimic an ATC’s priority assignment. Firstly, the potential conflict scenarios between two aircraft from historical data, which are resolved, are detected and extracted. Then a Random Forest model is developed to learn ATC’s behaviors. The model mimics ATC’s behavior with an accuracy of 89% and can thus be an effective approach for priority assignment in path-planning and conflict resolution. Further analysis indicates that features such as unimpeded time difference, distance to destination and start, and speed are major considerations that affect the ATC’s decisions. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

This study proposes an autonomous aircraft taxi-agent that can be used to recommend the pilot the optimal speed profile to achieve optimal fuel burn and to arrive on time at the target position on the taxiway while considering potential interactions with surrounding traffic. The problem is modeled as a control decision problem which is solved by training the agent under a Deep Reinforcement Learning (DRL) mechanism, using Proximal Policy Optimization (PPO) algorithm. The reward function is designed to consider the fuel burn, taxi-time, and delay-time. Thus, the trained agent will learn to taxi the aircraft between any pair of locations on the airport surface timely while maintaining safety and efficiency. As the result, in more than 97.8% of the evaluated sessions, the controlled aircraft can reach the target position with the time difference within the range of [-20,5] seconds. Moreover, compared with actual fuel burn, the proposed autonomous taxi-agent demonstrated a reduction of 29.5%, equivalent to the reduction of 13.9 kg of fuel per aircraft. This benefit in fuel burn reduction can complement the emission reductions achieved by solving other sub-problems, such as pushback control and taxi-route assignments to achieve much higher performance. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

Improving short-term air traffic flow prediction can help forecast demand and maximize existing capacity by tactical air traffic flow management. Most existing studies in flow prediction lacks consideration of the dynamic, structural, and interrelated nature of air traffic flows in the airspace. Therefore, this paper proposes to predict sector entry flows based on graph convolutional networks, which consider the dynamic spatial-temporal features of air traffic from a graph perspective. First, we specify a sector entry flow based on its upstream and downstream sectors. Then, each entry flow is denoted as a node in a graph. The weighted edges between the nodes are learned from a Word2vec model based on air traffic flows among the nodes. With the weighted graph constructed and the temporal flows on the nodes extracted from the flight trajectories, an Attention-based Spatial-Temporal Graph Convolutional Network (ASTGCN) module is adopted to capture spatial-temporal features of recent, daily-periodic, and weekly-periodic flows in the graph. Finally, The outputs from the ASTGCN module based on the three features are fused to generate the final prediction results. The proposed method is applied on 164 sectors of French airspace for one-month ADS-B data )from Dec 1, 2019, to Dec 31, 2019) which includes 158,856 flights. Results show that, the proposed method outperforms the well established Long short-term memory (LSTM) model, and demonstrates better capability in predicting rapid changes in traffic flow and has relatively smaller decrease in prediction accuracy as the prediction time-window increases. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

In drone operations, changes in the weather can potentially influence the planned routes and schedules of drones. It is therefore vital to incorporate proper models of weather uncertainty into drone flow management methods, such as the dynamic carousel circuit. As an extension of our previous studies, this research aims to monitor weather uncertainty and validate the efficiency and effectiveness of the dynamic carousel circuit when considering weather uncertainty and dynamical incoming flow. A comparison of prediction accuracies for first-order and second-order Markov chain models with simple weather states or realistic weather states is presented in this paper. Besides, a novel approach, two-layer simulation optimization, is introduced to solve the optimization problem for large-scale stochastic simulation efficiently. This proposed approach is composed of a segmental simulation optimization algorithm with a Genetic Algorithm to obtain the optimum radius for the dynamic carousel circuit of each interval parallelly, and a ranking and selection process to find the best candidate for a whole-day simulation duration among these optimum results efficiently. The finding of this study shows that such approach can be successfully applied to obtain the optimum radius for the dynamic carousel circuit with stochastic inputs. Results from the Monte Carlo simulation prove the stability of the dynamic carousel circuit under weather uncertainty and changing demand. Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University Submitted/Accepted version This research is supported by the Civil Aviation Authority of Singapore and Nanyang Technological University, Singapore under their collaboration in the Air Traffic Management Research Institute. This work was also supported by the National Natural Science Foundation of China (grant number 52102453) and Chongqing Research Program of Basic Research and Frontier Technology, China (grant number cstc2020jcyj-msxmX0602).

As applied to dynamic monitoring situations, the prevailing approach to situation awareness (SA) assumes that the relevant areas of interest (AOI) be perceived before that information can be processed further to effect decision making and thereafter action. It is not entirely clear whether this is the case. This study seeks to investigate the monitoring of dynamic objects through matching eye fixations with the relevant AOIs in boundary crossing scenarios. By this definition, a match is where a fixation is registered on the AOI. While many factors may affect monitoring characteristics, traffic simulations were designed in this study to explore two factors namely: the number of inbound/outbound traffic transfer and the number of entry and/or exit points in a radar monitoring sector. These two factors were graded into 5 levels of difficulty ranging from low to high traffic flow numbers. Combined permutation in terms of levels of difficulty of these two factors yielded a total of thirty scenarios. Through this, results showed that changes in the traffic flow numbers on transfer resulted in greater variations having match limits ranging from 29%-100%, as compared to the number of sector entry/exit points of range limit from 80%-100%. Subsequent analysis is able to determine the type and combination of traffic scenarios where imperfect matching is likely to occur. Civil Aviation Authority of Singapore (CAAS) Nanyang Technological University Published version This research is supported by the Civil Aviation Authority of Singapore and Nanyang Technological University, Singapore under their collaboration in the Air Traffic Management Research Institute.

Unmanned aerial vehicles (UAV) or ’drones’ are expected to increase in the future and be employed for a multitude of applications like parcel delivery, industrial inspection, aerial photography, and safety surveillance. Consequently, increase in UAV traffic may lead to higher likelihood of UAV failure and crash to occur. For urbanized environments, operation of UAV poses high risk to the population on ground due to increased population movement and expectation of people to be outdoors. Diffusive model simulation using random walk was employed to estimate population movement outside buildings at different times especially near Mass Rapid Transit (MRT) train stations. In our preliminary study, the influence of different times to the population movement was highlighted. It was observed that the population was more concentrated around MRT stations at morning and afternoon during weekdays, while more constant throughout the day during weekends or holidays. Ground risk mapping due to UAV operation could also be studied and performed in future study to have better understanding on this risk. Study of population movement in urbanized environments for the ground risk assessment of UAV operations will provide better insights on the limitation of current risk assessment. At the same time, this could help provide information for national aviation authorities to formulate risk assessment and approve UAV operation in urban environments. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

With the fast development of unmanned aircraft vehicle (UAV) operations in urban environments, the low-altitude air traffic is projected to be crowded. That would bring about problems such as how to determine the safe separation between two UAV operations. This paper proposes an evaluation framework for the determination of flight technical error (FTE) by investigating flight trajectory deviations, which is to support the establishment of separation requirement for UAV operations. Flight experiments are conducted by using a commonly used drone to collect flight trajectory data. Statistical analysis methods are used to investigate the trajectory deviations on lateral, longitudinal, and vertical axes. The normality tests are first conducted for the trajectory deviation data. The analysis of variance (ANOVA) is then performed to test the homogeneity of sample variances, while the regression analysis is conducted to investigate potential trends of the trajectory deviations with the change of flight heights. Statistical testing results show that the mean trajectory deviation in different flight times has no significant difference, whereas the mean trajectory deviation increases with the increase of flight height. The results also show that the linear increasing trend is more evident for deviations on the lateral track, while the quadratic increasing trend of trajectory deviation dominates on the longitudinal track. The deviations on vertical track is not significant compared with the ones on lateral and longitudinal tracks. This study investigates the flight technical error by analyzing the flight trajectory deviations, which will support the determination of separation requirement for safe UAV operations in low-altitude urban airspace. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Submitted/Accepted version This research is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme.

In an airway network, some critical links exist that are vital for the structural integrity and performance of the network. The detection of such links may assist with improving the imbalance between the limited airspace capacity and the ever-increasing traffic demand, which elicit flight delays, significant economic losses, etc. However, it is challenging to identify such links as they evolve (both in space and time) with changing traffic flow dynamics. This paper proposes a complex network approach for spatial-temporal critical links detection in a given airway network. First, flight track data is employed to characterize the airway network as weighted spatial-temporal networks. Then edge centrality and network percolation metrics are adopted to detect the critical links in each snapshot of the spatial-temporal networks. Afterward, the critical links detected by the two metrics are spatially overlapped to determine the final critical links over time. To examine the operational validity of the proposed method, we carry out a case study on the Southeast Asia airway network derived from one-month flight track data. Results demonstrate that the spatial distribution of the critical links varies over different traffic scenarios, and most of the identified critical links are found in the transition sectors with complex traffic situations. Four links, which are parts and at crosses of major trunk airways connecting to major navigation aids (VOR/DME) in the studied network, appear highly in all examined traffic scenarios. The unavailability of such links may lead to traffic flow disruptions. Observations by subject matter experts from air-traffic data visualizations demonstrate that the complex network-based methods can dynamically identify airway links that are operationally critical under time-evolving air-traffic scenarios. With good traffic flow prediction tools in the future, this method can be adopted to predict critical links in airway networks to better assist controllers in real-time air traffic management. Civil Aviation Authority of Singapore (CAAS) National Research Foundation (NRF) Published version This research / project* is supported by the National Research Foundation, Singapore, and the Civil Aviation Authority of Singapore, under the Aviation Transformation Programme. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of National Research Foundation, Singapore and the Civil Aviation Authority of Singapore.