With the increasing integration of Autonomous Vehicles (AVs) into traffic systems, the interaction between different types of vehicles presents a significant challenge for cooperative decision-making, particularly in mixed traffic environments where AVs coexist with Human-driven Vehicles (HVs). Among various scenarios, highway on-ramp merging is a critical and complex one where effective interaction is essential for ensuring traffic safety and efficiency, and decisions of AVs and HVs should be made efficiently. In this paper, we propose a Game-Theoretic Reinforcement Learning (GTRL)-based vehicle behavior interaction framework designed for various merging scenarios in mixed traffic environments. This framework includes Stackelberg leader-follower interactions between AVs and HVs, as well as Nash cooperative interactions between vehicles of the same type, such as AV-AV or HV-HV. We formulate the problem as a Partially Observable Markov Decision Process (POMDP) and employ the Bi-level Reinforcement Learning (Bi-RL) method and Safe Multi-Agent Deep Q-Network (MADQN) method to address two distinct types of interaction problems. A gym-based simulation environment is developed to evaluate four interaction scenarios between vehicles in mixed traffic. Comprehensive experimental results demonstrate the potential of the GTRL-based behavior interaction framework to adapt to the dynamic and uncertain nature of vehicle interactions.

Last-mile transportation, a critical component of public transit within integrated urban mobility systems, plays a pivotal role in promoting sustainability and requires immediate attention. However, due to the high real-time variability and uncertainties of last-mile travel demand, high passenger concurrency, and dispersed destinations, existing last-mile transportation systems often encounter significant challenges. These challenges include resource shortages and congestion during peak hours, as well as high operational costs and excessive passenger waiting times during off-peak periods. With the rapid development and widespread adoption of autonomous vehicles, which are characterized by centralized control and flexible scheduling, this study proposes leveraging heterogeneous autonomous vehicles to address these challenges. Specifically, a mixed-integer programming model is developed to maximize the service provider’s profit by considering fare profit, passenger waiting time penalties, and operating costs. To enable real-time decision-making, then an attention-based deep reinforcement learning algorithm is introduced. This algorithm incorporates two decoder mechanisms for vehicle selection and passenger allocation in the scheduling of heterogeneous autonomous vehicles. This involves dynamically selecting vehicles from a heterogeneous fleet based on passenger demand using an attention mechanism, optimizing efficiency in serving last-mile travelers. Extensive numerical experiments and a real-world case study across various datasets demonstrate that the proposed service model and algorithm effectively solve the scheduling problem while meeting the demands of on-demand last-mile transportation. Furthermore, these innovations contribute to reducing fleet carbon emissions and advancing sustainable urban transportation.

Emergency medical delivery plays a crucial role in ensuring timely treatment for patients under the promising trend of medical resource sharing. However, traditional human courier based delivery modes face many problems, such as high costs, slow speeds, and uncertain arrival times due to traffic conditions. To address these issues, this paper proposes leveraging heterogeneous Unmanned Aerial Vehicles (UAVs) and human couriers for emergency medical delivery, and develops a Cooperative Deep Reinforcement Learning (DRL) based method for real-time joint scheduling. The problem is modelled as a multi-depot capacitated pickup and delivery problem with soft deadlines. A DRL-based method is proposed with two types of agent networks for UAVs and human couriers, respectively, which could capture their distinct features and delivery strategies. In addition, a cooperative network is introduced to coordinate their operations. Extensive computational experiments and a real-life case study are conducted that verifies the superiority of our methods over several benchmark algorithms in different scenarios, and demonstrates its feasibility and performance in practical scenarios.

Visual detection failures are a critical challenge in air traffic control (ATC), where undetected alerts can compromise operational safety and decision-making. Previous studies have primarily assessed detection failures through target fixation patterns, yet this method struggles to identify the more complex “look-but-fail-to-see” and “see-without-looking” scenarios. This underscores the necessity of exploring peripheral vision mechanisms, where dynamic tracking trajectories could better capture the scope of visual attention. Therefore, this study proposes a classification framework for visual detection by integrating peripheral vision tracking and human attentional states, including detection failures such as peripheral vision neglect and look-but-fail-to-see errors. A hierarchical detection failure recognition framework specific to the ATC settings is further developed and validated through an ATC simulation experiment. The framework first employs an Adaptive Symbolic Alert Detection method to identify and annotate ATC-specific alert regions with spatiotemporal uncertainty (achieving 95.24% precision), followed by LLM-based evaluation of operators’ visual attention to these regions to intelligently assign classification labels. Additionally, we introduce a fixation coordinate-sensitive multi-domain feature set that captures spatiotemporal and frequency-domain characteristics across detection types, achieving 93.13% four-class classification accuracy, outperforming traditional feature sets (83.69%) and both single-and dual-domain features (ranging from 76.82% to 90.11% accuracy). These findings demonstrate that our framework effectively captures a broader and structured range of visual detection failures, providing critical insights to improve the reliability of alert detection in ATC and the design of an intelligent human-centered ATC support system.

Unmanned Aerial Vehicle (UAV) has been increasingly adopted for IoT data collection in large-scale scenarios that are with less or even no network coverage. Efficient UAV route planning is a vital part of such a UAV-based data collection process, which is recognized to be complex and challenging, especially considering that the amount of data collected is dependent on UAV visit time and service time and many coupled decisions are involved. Taking these challenges into consideration, this paper proposes a new hybrid heuristics-based UAV route planning method for IoT data collection. Specifically, the relationships among UAV service time, data amount, and data collection time windows of IoT devices are analyzed first, then an integrated route planning model for multiple UAVs is developed. After that, an innovative Hybrid Tabu Search-Variable Neighborhood Descent (HTS-VND) algorithm is developed, with six effective operators that could further improve computing efficiency and solution quality. Finally, extensive experimental case studies are conducted. The proposed method can efficiently improve the collected data amount compared to existing methods in medium-scale and large-scale scenarios.

Accurate fault diagnosis of rotating machinery is essential for smooth and safe operations of mechanical systems, and various data-driven methods have been developed based on massive sensing data. However, the frequent occurrence of compound faults makes it much challenging. Meanwhile, the few labeled and imbalanced data of rotating machinery further complicate the design of diagnosis methods. To address these issues, this article proposes a novel sequential feature augmented deep multilabel learning model for compound fault diagnosis. Specifically, by integrating convolutional neural network with convolutional long short-term memory, a deep stacked sparse autoencoder is developed to extract high-dimensional marginal and time-sequential features from few labeled and imbalanced data. Then, a supervised multilabel learning model is developed to learn the relationships among features of single and compound faults and finally realize accurate compound fault diagnosis. Experimental results demonstrated that our model could cope well with few labeled and imbalanced data scenarios and outperforms many existing models.

The congestion and disruption of information infrastructures frequently happen during disasters, which would hinder the understanding of disaster scenarios, and thus impede rapid response activities. With the advantages of high flexibility and efficiency, this paper proposes to use UAVs as temporary and mobile relays for disaster data collection. However, different from many existing data collection scenarios in industrial sectors, the disaster data value varies with UAV arrival time and service time in terms of their importance for disaster response, which makes the scheduling of UAVs challenging. To address such a problem, this paper proposes an attention-based Deep Reinforcement Learning (DRL) method for multi-UAV scheduling considering time-varying data value. Specifically, the problem is modeled as a specific team orienteering problem with time-varying value. Then the relationships between UAV route selection and service time at each node are analyzed, based on which the computing efficiency for solution algorithms can be improved. After that, an attention-based DRL method is developed, with a calibrated attention model and decoding method. Finally, systematic computational experiments are conducted to evaluate the performance of the proposed method, which demonstrates its superiority over popular methods in UAV scheduling, especially for large-scale and complex scenarios.

Airport baggage transport is a key aspect of airport ground handling which involves transporting passenger baggage between the baggage handling center and aircraft stands. Typically, tractors and trailers are used to implement the baggage transport service under the drop-and-pull mode in practice. Efficient airport baggage transport service is essential to reduce flight delays, save airport ground handling costs, and guarantee aviation safety. However, the existing methods for scheduling tractors and trailers perform poorly when dealing with high flight volumes at busy hub airports, leading to delayed responses for baggage transport demands. Besides, few works have explored the significance of the drop-and-pull mode in improving airport baggage transport service equality. Thus, to enhance the efficiency of airport baggage transport service and reduce airport ground handling costs, this paper presents an integrated tractor and trailer scheduling problem under the drop-and-pull mode. Then, an improved Genetic Algorithm (GA) is developed to solve this problem. Finally, simulation experiments are conducted to verify the effectiveness of the model and algorithm.

The adoption of robotics is promising to improve the efficiency, quality, and safety of prefabricated construction. Besides technologies that improve the capability of a single robot, the automated assembly planning for robots at construction sites is vital for further improving the efficiency and promoting robots into practices. However, considering the highly dynamic and uncertain nature of a construction environment, and the varied scenarios in different construction sites, it is always challenging to make appropriate and up-to-date assembly plans. Therefore, this paper proposes a Deep Reinforcement Learning (DRL) based method for automated assembly planning in robot-based prefabricated construction. Specifically, a re-configurable simulator for assembly planning is developed based on a Building Information Model (BIM) and an open game engine, which could support the training and testing of various optimization methods. Furthermore, the assembly planning problem is modelled as a Markov Decision Process (MDP) and a set of DRL algorithms are developed and trained using the simulator. Finally, experimental case studies in four typical scenarios are conducted, and the performance of our proposed methods have been verified, which can also serve as benchmarks for future research works within the community of automated construction.Note to Practitioners—This paper is conducted based on the comprehensive analysis of real-life assembly planning processes in prefabricated construction, and the methods proposed could bring many benefits to practitioners. Firstly, the proposed simulator could be easily re-configured to simulate diverse scenarios, which can be used to evaluate and verify the operations’ optimization methods and new construction technologies. Secondly, the proposed DRL-based optimization methods can be directly adopted in various robot-based construction scenarios, and can also be tailored to support the assembly planning in traditional human-based or human-robot construction environments. Thirdly, the proposed DRL methods and their performance in the four typical scenarios can serve as benchmarks for proposing new advanced construction technologies and optimization methods in assembly planning.

There is a growing number of industrial parks implementing supply hubs to provide shared logistics services for member manufacturers. Products from manufacturers could be temporarily stored at the supply hub in industrial park (SHIP), and consolidated into one or more vehicles to be delivered to their retail distribution centers. This article discusses how SHIP and heterogeneous manufacturers interact to optimize their decisions on consolidated transportation service pricing and consolidation schedules. An all-unit multiple-breakpoint quantity discount pricing scheme is adopted. This problem is modeled as a bilevel program where the SHIP is treated as the leader and manufacturers as followers. Optimal properties of the proposed model are analyzed and the global optimal solution is derived. Numerical studies are conducted to examine the effectiveness of this pricing scheme through comparing with the regular transportation price under various circumstances. The results show that the SHIP operator could always increase profit by using this pricing scheme, while manufacturers’ performances are not affected. Among heterogeneous manufacturers with various production capabilities and resources, the SHIP could gain the largest profit increase through adopting quantity discount pricing from medium-sized manufacturers.

In this article, we study the reinforcement learning (RL) for vehicle routing problems (VRPs). Recent works have shown that attention-based RL models outperform recurrent neural network-based methods on these problems in terms of both effectiveness and efficiency. However, existing RL models simply aggregate node embeddings to generate the context embedding without taking into account the dynamic network structures, making them incapable of modeling the state transition and action selection dynamics. In this work, we develop a new attention-based RL model that provides enhanced node embeddings via batch normalization reordering and gate aggregation, as well as dynamic-aware context embedding through an attentive aggregation module on multiple relational structures. We conduct experiments on five types of VRPs: 1) travelling salesman problem (TSP); 2) capacitated VRP (CVRP); 3) split delivery VRP (SDVRP); 4) orienteering problem (OP); and 5) prize collecting TSP (PCTSP). The results show that our model not only outperforms the learning-based baselines but also solves the problems much faster than the traditional baselines. In addition, our model shows improved generalizability when being evaluated in large-scale problems, as well as problems with different data distributions.

Nowadays, increasing attention has been drawn to hybrid 2D-3D visualization tools, while evaluating them with a convenient and objective tool has only been carried out in a small number of areas. In this study, a revised radar chart-based usability evaluation approach was proposed. The approach was adopted to evaluate the hybrid 2D-3D radar display in air traffic management. The holding stack in air traffic management is analyzed and simulated, two generic tasks are designed accordingly. The hybrid 2D-3D radar display settings are evaluated based on six indicators from eye-tracking and brain dynamics data, namely, the frequency of fixation, fixation mean duration, fixation time on an area of interest, emotion, workload, and stress. The results reveal that the hybrid 2D-3D radar display induces spatial memory loss and high workload, while requires a shorter fixation duration.

Objectively measuring and monitoring human mental states in a non-intrusive way is important in improving the context-awareness of smart objects. One of the suitable bio-signals in measuring human mental states is aye-tracking data, as visual is the first channel of information collection. In addition, eye-tracking data shows the process of human-system interactions. Traditionally, many studies have been conducted to investigate the correlations between eye-tracking data and human mental states. Recently, with advanced artificial intelligence algorithms, the spatial and temporal patterns of eye-tracking data can be deeply analyzed for detecting human mental states. This study aims to explore and review eye-tracking parameters and state-of-art methods for mental states assessments. The study reveals that both statistical methods and novel methods, such as machine learning and deep learning have been applied to process eye-tracking data. Besides, novel features extracted from eye-tracking data, such as gaze-bin and entropy have been used in assessing human mental states. This review is expected to provide references for eye-tracking data analysis.

Eye-tracking-based human fatigue detection at traffic control centers suffers from an unavoidable problem of low-quality eye-tracking data caused by noisy and missing gaze points. In this article, the authors conducted pioneering work by investigating the effects of data quality on eye-tracking-based fatigue indicators and by proposing a hierarchical-based interpolation approach to extract the eye-tracking-based fatigue indicators from low-quality eye-tracking data. This approach adaptively classified the missing gaze points and hierarchically interpolated them based on the temporal-spatial characteristics of the gaze points. In addition, the definitions of applicable fixations and saccades for human fatigue detection is proposed. Two experiments are conducted to verify the effectiveness and efficiency of the method in extracting eye-tracking-based fatigue indicators and detecting human fatigue. The results indicate that most eye-tracking parameters are significantly affected by the quality of the eye-tracking data. In addition, the proposed approach can achieve much better performance than the classic velocity threshold identification algorithm (I-VT) and a state-of-the-art method (U'n'Eye) in parsing low-quality eye-tracking data. Specifically, the proposed method attained relatively stable eye-tracking-based fatigue indicators and reported the highest accuracy in human fatigue detection. These results are expected to facilitate the application of eye movement-based human fatigue detection in practice.

Emergency fleet management has become one of the determinant success factors for post-disaster responses in urban flood control. However, it is challenging as multiple types of emergency vehicles are involved, and its performance is frequently threatened by the fluctuation of rescue demands and fleet capacity. Aiming at coping with the imbalances between rescue demands and vehicle supplies, and maintaining required service level of fleet management after flood occurs, this paper proposes a data-driven resilient fleet management solution under the context of cloud asset-enabled urban flood control. First, the problem of resilient fleet management is quantitatively defined, and then a data-driven dynamic management mechanism is proposed, which is highly effective on realizing resilient fleet management. Furthermore, considering the cooperation among different types of emergency vehicles, a greedy-based algorithm is proposed for resilient vehicle dispatching based on real-time scenarios. Finally, a simulation case is also conducted to verify the effectiveness and performance of the proposed solution.

Urban flood control has become an important issue for both city planners and researchers as flood is one of the most severe natural disasters for large cities. Although many measures have been implemented for urban flood control, the effectiveness and efficiency is greatly hampered by the poor management of physical assets. In urban flood control, managing physical assets is challenging as it involves a complex asset base with the dispersion of asset allocations, management processes, and information systems. To address these challenges, this paper proposes an integrated asset management platform for urban flood control. Cloud technology is adopted to provide whole life cycle management of physical assets by integrating various management processes and information systems. It also enables the cooperative management between sectors. Besides, the concept of smart asset is introduced to realize remote management and real-time data collection from these assets. In the end, a case study is also given to demonstrate the working procedural of managing a typical smart asset through the proposed cloud platform.