Yutian Pang

Yutian Pang, Jueming Hu, Christopher S. Lieber et al.

Air traffic control (ATC) is a safety-critical service system that demands constant attention from ground air traffic controllers (ATCos) to maintain daily aviation operations. The workload of the ATCos can have negative effects on operational safety and airspace usage. To avoid overloading and ensure an acceptable workload level for the ATCos, it is important to predict the ATCos' workload accurately for mitigation actions. In this paper, we first perform a review of research on ATCo workload, mostly from the air traffic perspective. Then, we briefly introduce the setup of the human-in-the-loop (HITL) simulations with retired ATCos, where the air traffic data and workload labels are obtained. The simulations are conducted under three Phoenix approach scenarios while the human ATCos are requested to self-evaluate their workload ratings (i.e., low-1 to high-7). Preliminary data analysis is conducted. Next, we propose a graph-based deep-learning framework with conformal prediction to identify the ATCo workload levels. The number of aircraft under the controller's control varies both spatially and temporally, resulting in dynamically evolving graphs. The experiment results suggest that (a) besides the traffic density feature, the traffic conflict feature contributes to the workload prediction capabilities (i.e., minimum horizontal/vertical separation distance); (b) directly learning from the spatiotemporal graph layout of airspace with graph neural network can achieve higher prediction accuracy, compare to hand-crafted traffic complexity features; (c) conformal prediction is a valuable tool to further boost model prediction accuracy, resulting a range of predicted workload labels. The code used is available at \href{https://github.com/ymlasu/para-atm-collection/blob/master/air-traffic-prediction/ATC-Workload-Prediction/}{$\mathsf{Link}$}.

Rahul Rathnakumar, Yutian Pang, Yongming Liu

Accurately detecting crack boundaries is crucial for reliability assessment and risk management of structures and materials, such as structural health monitoring, diagnostics, prognostics, and maintenance scheduling. Uncertainty quantification of crack detection is challenging due to various stochastic factors, such as measurement noises, signal processing, and model simplifications. A machine learning-based approach is proposed to quantify both epistemic and aleatoric uncertainties concurrently. We introduce a Bayesian Boundary-Aware Convolutional Network (B-BACN) that emphasizes uncertainty-aware boundary refinement to generate precise and reliable crack boundary detections. The proposed method employs a multi-task learning approach, where we use Monte Carlo Dropout to learn the epistemic uncertainty and a Gaussian sampling function to predict each sample's aleatoric uncertainty. Moreover, we include a boundary refinement loss to B-BACN to enhance the determination of defect boundaries. The proposed method is demonstrated with benchmark experimental results and compared with several existing methods. The experimental results illustrate the effectiveness of our proposed approach in uncertainty-aware crack boundary detection, minimizing misclassification rate, and improving model calibration capabilities.

Jiayu Huang, Yutian Pang, Yongming Liu et al.

Neural Networks (NNs) have been widely {used in supervised learning} due to their ability to model complex nonlinear patterns, often presented in high-dimensional data such as images and text. However, traditional NNs often lack the ability for uncertainty quantification. Bayesian NNs (BNNS) could help measure the uncertainty by considering the distributions of the NN model parameters. Besides, domain knowledge is commonly available and could improve the performance of BNNs if it can be appropriately incorporated. In this work, we propose a novel Posterior-Regularized Bayesian Neural Network (PR-BNN) model by incorporating different types of knowledge constraints, such as the soft and hard constraints, as a posterior regularization term. Furthermore, we propose to combine the augmented Lagrangian method and the existing BNN solvers for efficient inference. The experiments in simulation and two case studies about aviation landing prediction and solar energy output prediction have shown the knowledge constraints and the performance improvement of the proposed model over traditional BNNs without the constraints.

Yutian Pang, Andrew Paul Kendall, John-Paul Clarke

Mission-critical operations of highly maneuverable Remotely Piloted Aircraft Systems (RPAS) require reliable communication to ensure safe integration into existing airspace. Understanding system-level performance under stochastic communication conditions is essential for estimating mission success and assessing operational risks. This study quantifies the impact of communication latency and complete signal loss on the mission completion performance of a highly maneuverable RPAS. The mission is defined as a static waypoint tracking task in three-dimensional airspace. We first derive mathematical formulations for key reliability metrics within the Required Communication Performance (RCP) framework. These stochastic communication factors, including latency and availability, are then incorporated into flight control simulations to evaluate system behavior. Extensive multiprocessing Monte Carlo simulations are conducted using high-performance computing to generate mission success rate and mission completion time envelopes. Results show significant degradation in flight performance as communication latency increases or availability decreases, which directly reduces the system stability margin. To better characterize this relationship, we introduce a new reliability metric, communicability, which integrates three key RCP metrics and provides insight into the maximum tolerable latency for flight control. The proposed framework informs RPAS design by revealing trade-offs between communication capability and flight control performance. The code used in this study is publicly available at this \href{https://github.com/YutianPangASU/comm-dynamics}{repository}.

Yutian Pang, Peng Zhao, Jueming Hu et al.

This paper addresses aircraft delays, emphasizing their impact on safety and financial losses. To mitigate these issues, an innovative machine learning (ML)-enhanced landing scheduling methodology is proposed, aiming to improve automation and safety. Analyzing flight arrival delay scenarios reveals strong multimodal distributions and clusters in arrival flight time durations. A multi-stage conditional ML predictor enhances separation time prediction based on flight events. ML predictions are then integrated as safety constraints in a time-constrained traveling salesman problem formulation, solved using mixed-integer linear programming (MILP). Historical flight recordings and model predictions address uncertainties between successive flights, ensuring reliability. The proposed method is validated using real-world data from the Atlanta Air Route Traffic Control Center (ARTCC ZTL). Case studies demonstrate an average 17.2% reduction in total landing time compared to the First-Come-First-Served (FCFS) rule. Unlike FCFS, the proposed methodology considers uncertainties, instilling confidence in scheduling. The study concludes with remarks and outlines future research directions.

Yutian Pang, Zehua Guo, Binnan Zhuang

Behavior prediction plays an important role in integrated autonomous driving software solutions. In behavior prediction research, interactive behavior prediction is a less-explored area, compared to single-agent behavior prediction. Predicting the motion of interactive agents requires initiating novel mechanisms to capture the joint behaviors of the interactive pairs. In this work, we formulate the end-to-end joint prediction problem as a sequential learning process of marginal learning and joint learning of vehicle behaviors. We propose ProspectNet, a joint learning block that adopts the weighted attention score to model the mutual influence between interactive agent pairs. The joint learning block first weighs the multi-modal predicted candidate trajectories, then updates the ego-agent's embedding via cross attention. Furthermore, we broadcast the individual future predictions for each interactive agent into a pair-wise scoring module to select the top $K$ prediction pairs. We show that ProspectNet outperforms the Cartesian product of two marginal predictions, and achieves comparable performance on the Waymo Interactive Motion Prediction benchmarks.

Yutian Pang, Daniel Delahaye, John-Paul Clarke

We present a trajectory-based optimization framework for arrival sequencing and scheduling in the terminal maneuvering area (TMA). Unlike node-link scheduling models that reduce trajectories to time-delay variables, the proposed method computes implementable per-aircraft speed profiles and path extensions that achieve required landing separation through terminal air traffic control actions. The framework combines an analytic TMA path model, consisting of a tangent leg, a radius-to-fix turn, and a final-approach segment, with a nonlinear program (NLP) that jointly optimizes path stretch and segment speeds under a weighted objective. Three landing-order policies are examined: First-Entry-First-Serve (FEFS), First-on-Final-First-Serve (FOFFS), and FOFFS with Constrained Position Shifting (CPS) up to $k$ positions. CPS is implemented through a two-phase approach coupling mixed-integer linear programming (MILP) with NLP to select an optimized landing order before trajectory optimization. The aircraft population follows a realistic weight-class fleet mix with pair-specific wake-turbulence separation, and each scenario is perturbed by a Gaussian wind sample projected onto each segment to convert commanded airspeeds into ground speeds. An online rolling-horizon formulation commits each aircraft trajectory irrevocably upon entry, enabling real-time decision-making. Monte Carlo experiments on the simplified A80 TMA show that: (i) FOFFS consistently outperforms FEFS in delay, path stretch, and fuel burn by exploiting geometric asymmetries among arrival streams; (ii) CPS further reduces separation violations and path stretch, though with diminishing returns and rapidly increasing solver cost; (iii) fuel estimates from BADA 3 and OpenAP show consistent qualitative trends; and (iv) per-entry optimization completes in near real-time, supporting practical deployment.

Yutian Pang, Andrew Kendall, John-Paul Clarke

We investigate the potential impact of communication and human performance uncertainties on runway operations. Specifically, we consider these impacts within the context of an arrival scenario with two converging flows: a straight-in approach stream and a downwind stream merging into it. Both arrival stream are modeled using a modified Possion distribution that incorporate the separation minima as well as the runway occupancy time. Various system level uncertainties are addressed in this process, including communication link- and human-related uncertainties. In this research, we first build a Monte Carlo-based discrete-time simulation, where aircraft arrivals are generated by modified Poisson processes subject to minimum separation constraints, simulating various traffic operations. The merging logic incorporates standard bank angle continuous turn-to-final, pilot response delays, and dynamic gap availability in real time. Then, we investigate an automated final approach vectoring model (i.e., Auto-ATC), in which inverse optimal control is used to learn decision advisories from human expert records. By augmenting trajectories and incorporating the aforementioned uncertainties into the planning scenario, we create a setup analogous to the discrete event simulation. For both studies, runway capacity is measured by runway throughput, the fraction of downwind arrivals that merge immediately without holding, and the average delay (i.e., holding time/distance) experienced on the downwind leg. This research provides a method for runway capacity estimation in merging scenarios, and demonstrates that aeronautical communication link uncertainties significantly affect runway capacity in current voice-based operations, whereas the impact can be mitigated in autonomous operational settings.

Jueming Hu, Zhe Xu, Weichang Wang et al.

In multi-agent reinforcement learning (MARL), it is challenging for a collection of agents to learn complex temporally extended tasks. The difficulties lie in computational complexity and how to learn the high-level ideas behind reward functions. We study the graph-based Markov Decision Process (MDP) where the dynamics of neighboring agents are coupled. We use a reward machine (RM) to encode each agent's task and expose reward function internal structures. RM has the capacity to describe high-level knowledge and encode non-Markovian reward functions. We propose a decentralized learning algorithm to tackle computational complexity, called decentralized graph-based reinforcement learning using reward machines (DGRM), that equips each agent with a localized policy, allowing agents to make decisions independently, based on the information available to the agents. DGRM uses the actor-critic structure, and we introduce the tabular Q-function for discrete state problems. We show that the dependency of Q-function on other agents decreases exponentially as the distance between them increases. Furthermore, the complexity of DGRM is related to the local information size of the largest $κ$-hop neighborhood, and DGRM can find an $O(ρ^{κ+1})$-approximation of a stationary point of the objective function. To further improve efficiency, we also propose the deep DGRM algorithm, using deep neural networks to approximate the Q-function and policy function to solve large-scale or continuous state problems. The effectiveness of the proposed DGRM algorithm is evaluated by two case studies, UAV package delivery and COVID-19 pandemic mitigation. Experimental results show that local information is sufficient for DGRM and agents can accomplish complex tasks with the help of RM. DGRM improves the global accumulated reward by 119% compared to the baseline in the case of COVID-19 pandemic mitigation.

Yutian Pang, Sheng Cheng, Jueming Hu et al.

To evaluate the robustness gain of Bayesian neural networks on image classification tasks, we perform input perturbations, and adversarial attacks to the state-of-the-art Bayesian neural networks, with a benchmark CNN model as reference. The attacks are selected to simulate signal interference and cyberattacks towards CNN-based machine learning systems. The result shows that a Bayesian neural network achieves significantly higher robustness against adversarial attacks generated against a deterministic neural network model, without adversarial training. The Bayesian posterior can act as the safety precursor of ongoing malicious activities. Furthermore, we show that the stochastic classifier after the deterministic CNN extractor has sufficient robustness enhancement rather than a stochastic feature extractor before the stochastic classifier. This advises on utilizing stochastic layers in building decision-making pipelines within a safety-critical domain.