Traffic accidents result in millions of injuries and fatalities globally, with a significant number occurring at intersections each year. Traffic Signal Control (TSC) is an effective strategy for enhancing safety at these urban junctures. Despite the growing popularity of Reinforcement Learning (RL) methods in optimizing TSC, these methods often prioritize driving efficiency over safety, thus failing to address the critical balance between these two aspects. Additionally, these methods usually need more interpretability. CounterFactual (CF) learning is a promising approach for various causal analysis fields. In this study, we introduce a novel framework to improve RL for safety aspects in TSC. This framework introduces a novel method based on CF learning to address the question: ``What if, when an unsafe event occurs, we backtrack to perform alternative actions, and will this unsafe event still occur in the subsequent period?'' To answer this question, we propose a new structure causal model to predict the result after executing different actions, and we propose a new CF module that integrates with additional ``X'' modules to promote safe RL practices. Our new algorithm, CFLight, which is derived from this framework, effectively tackles challenging safety events and significantly improves safety at intersections through a near-zero collision control strategy. Through extensive numerical experiments on both real-world and synthetic datasets, we demonstrate that CFLight reduces collisions and improves overall traffic performance compared to conventional RL methods and the recent safe RL model. Moreover, our method represents a generalized and safe framework for RL methods, opening possibilities for applications in other domains. The data and code are available in the github https://github.com/AdvancedAI-ComplexSystem/SmartCity/tree/main/CFLight.

Abstract With the refinement of 3D imaging modality, cryo-electron tomography (cryo-ET) has emerged as a powerful technique for the structural analysis of macromolecular complexes at near-atomic resolution. Recent advancements in volumetric segmentation methods applied to cryo-ET datasets have garnered significant attention within the biomedical sector. However, existing methods rely heavily on manually labeled data, which demands highly specialized expertise, making fully supervised approaches less feasible for cryo-ET images. To address this, a number of unsupervised domain adaptation (UDA) techniques have been developed to improve segmentation network performance using unlabeled data. Nevertheless, directly applying these methods to cryo-ET image segmentation presents two major challenges: 1) the source dataset, usually obtained through simulation, contains a fixed level of noise, while the target dataset, being directly collected from raw-data from the real-world scenario, has unpredictable noise levels; 2) the source data used for training typically consists of known macromolecules, in contrast, the target domain data are often unknown, causing the model to be biased towards those known macromolecules, leading to a domain shift problem. To address such challenges, in this paper, we introduce a voxel-wise unsupervised domain adaptation approach, termed Vox-UDA, specifically for cryo-ET subtomogram segmentation. Vox-UDA incorporates a noise generation module to simulate target-like noises in the source dataset for cross-noise level adaptation, and a denoised pseudo-labeling strategy based on the improved bilateral filter to alleviate the domain shift problem. Additionally, we further consider a scenario that is more in line with the real world, where the target (experimental) dataset might not be accessible during training or has a very small sample size, and we present a voxel-wise zero-shot domain adaptation (ZSDA) approach, named Vox-ZSDA. In Vox-ZSDA, we introduce a self-supervised graph learning strategy to eliminate any dependency on the target data, accompanied by a dynamic graph contrastive learning technique to enhance the model’s sensitivity to macromolecular structures to boost the segmentation performance. More importantly, we construct the first UDA and ZSDA cryo-ET subtomogram segmentation benchmark on three experimental datasets. Extensive experimental results on multiple benchmarks and newly curated real-world datasets demonstrate the superiority of our proposed approach compared to state-of-the-art UDA and ZSDA methods.

Accurate traffic speed prediction is a critical precursor for traffic management activities such as congestion prevention, accident warning, and traffic network optimization. Given the complexity of traffic flow data and the spatiotemporal correlations within this data, we present a Spatiotemporal Multi-head Graph Attention (STMGAT) traffic speed prediction model, which consists of multiple layers for extracting spatiotemporal features from traffic speed data. Each layer is constructed by serially connecting a Bidirectional Long Short-Term Memory Network (Bi-LSTM), a Multi-head Graph Attention Network (MGAT), and a gated causal Convolution Neural Network (gated causal CNN). In this structure, Bi-LSTM and gated causal CNN are responsible for two stages of temporal feature extraction from the traffic speed, while the MGAT, positioned between Bi-LSTM and gated causal CNN, facilitates the transfer of temporal features and the aggregation of spatial features. Furthermore, a Bayesian Optimization-Evolutionary Algorithm (BO-EA) mutual feedback optimization framework is proposed to optimize the network architecture and hyperparameters of the STMGAT model. The evolutionary algorithm provides evolutionarily-informed sampling points to the Bayesian optimization, while the Bayesian optimization supplies exogenous population individuals to the evolutionary algorithm. The two algorithms effectively optimize the network architecture and hyperparameters of the STMGAT model by continuously exchanging optimization information with each other. Experimental results demonstrate that the STMGAT model, featured with a decision-based and adaptive architecture, can effectively capture the complex spatiotemporal features in traffic speed data, achieving superior prediction performance compared to the baseline models. Moreover, the BO-EA mutual feedback framework shows higher efficiency on optimizing the network architecture and hyperparameters compared to single optimization algorithms.

Cryo-Electron Tomography (cryo-ET) is a 3D imaging technology that facilitates the study of macromolecular structures at near-atomic resolution. Recent volumetric segmentation approaches on cryo-ET images have drawn widespread interest in the biological sector. However, existing methods heavily rely on manually labeled data, which requires highly professional skills, thereby hindering the adoption of fully-supervised approaches for cryo-ET images. Some unsupervised domain adaptation (UDA) approaches have been designed to enhance the segmentation network performance using unlabeled data. However, applying these methods directly to cryo-ET image segmentation tasks remains challenging due to two main issues: 1) the source dataset, usually obtained through simulation, contains a fixed level of noise, while the target dataset, directly collected from raw-data from the real-world scenario, have unpredictable noise levels. 2) the source data used for training typically consists of known macromoleculars. In contrast, the target domain data are often unknown, causing the model to be biased towards those known macromolecules, leading to a domain shift problem. To address such challenges, in this work, we introduce a voxel-wise unsupervised domain adaptation approach, termed Vox-UDA, specifically for cryo-ET subtomogram segmentation. Vox-UDA incorporates a noise generation module to simulate target-like noises in the source dataset for cross-noise level adaptation. Additionally, we propose a denoised pseudo-labeling strategy based on the improved Bilateral Filter to alleviate the domain shift problem. More importantly, we construct the first UDA cryo-ET subtomogram segmentation benchmark on three experimental datasets. Extensive experimental results on multiple benchmarks and newly curated real-world datasets demonstrate the superiority of our proposed approach compared to state-of-the-art UDA methods.

U-Net architectures are widely used in skin lesion segmentation due to their ability to capture fine-grained spatial details and context. However, most models enhance performance by incorporating complex modules, often overlooking the computational resource constraints present in real-world medical environments. Consequently, there is an urgent need to design models for mobile skin lesion segmentation that are efficient in terms of both parameters and computational load. To address this issue, we propose a lightweight skin lesion segmentation network based on a differential evolution algorithm, the Lightweight Differential Evolution-based Boundary-Assisted UNet for Skin Lesion Segmentation (LDEB-UNet). LDEB-UNet introduces three main innovations: (1) We propose the Separable Convolution with GELU and SE module (SCGS), incorporated into Stages 1-3 of the U-shaped architecture. Utilizing depthwise separable convolutions effectively reduces the model's parameter count, while the addition of channel attention and GELU further enhances its performance. (2) We propose the Hybrid Group Attention Shuffle module (HGAS), which enhances feature interaction across different channel segments, improving model performance without increasing the parameter count. (3) We introduce a Differential Evolution-based Boundary-Assisted module(DEB) into segmentation networks, which improves the model's ability to handle blurry boundaries. Comprehensive experiments on the ISIC 2017 and ISIC 2018 datasets demonstrate that LDEB-UNet outperforms existing state-of-the-art methods. Moreover, to our best knowledge, this is the first model with a parameter count limited to just$\mathbf{2 6 K B}$and Giga-Operations Per Second (GFLOPs) limited to 0.081. Our code is available at https://github.com/cjr851/LDEB-UNet.

Effective traffic signal control (TSC) is crucial in mitigating urban congestion and reducing emissions. Recently, reinforcement learning (RL) has been the research trend for TSC. However, existing RL algorithms face several real-world challenges that hinder their practical deployment in TSC: (1) Sensor accuracy deteriorates with increased sensor detection range, and data transmission is prone to noise, potentially resulting in unsafe TSC decisions. (2) During the training of online RL, interactions with the environment could be unstable, potentially leading to inappropriate traffic signal phase (TSP) selection and traffic congestion. (3) Most current TSC algorithms focus only on TSP decisions, overlooking the critical aspect of phase duration, affecting safety and efficiency. To overcome these challenges, we propose a robust two-stage fuzzy approach called FuzzyLight, which integrates compressed sensing and RL for TSC deployment. FuzzyLight offers several key contributions: (1) It employs fuzzy logic and compressed sensing to address sensor noise and enhances the efficiency of TSP decisions. (2) It maintains stable performance during training and combines fuzzy logic with RL to generate precise phases. (3) It works in real cities across 22 intersections and demonstrates superior performance in both real-world and simulated environments. Experimental results indicate that FuzzyLight enhances traffic efficiency by 48% compared to expert-designed timings in the real world. Furthermore, it achieves state-of-the-art (SOTA) performance in simulated environments using six real-world datasets with transmission noise. The code and deployment video are available at the Github.

A significant threat to GPS users, including Unmanned Aerial Vehicles (UAVs), is the location spoofing attack, which can mislead systems with false GPS signals, jeopardising their operations and safety. To address this challenge, this study presents a method for verifying GPS spoofing attacks on UAV systems. The proposed solution develops a robust methodology by analysing the reported position of the UAV, along with various features of the received signal, such as the signal-to-noise ratio (SNR), azimuth, and pitch angles, at multiple base stations. Additionally, we consider Nakagami fading channels to model the properties of the received signal, which are relevant to real-world scenarios. We developed a smart verification algorithm using the Recurrent Neural Network (RNN) to authenticate the position reported by UAVs based on SNR, azimuth, and pitch angle data at various base station antennas. We have utilised both simple recurrent neural network (SRNN) and long short-term memory (LSTM) algorithms to evaluate and compare the performance of the model by varying the number of base stations. The performance of the algorithm was evaluated using confusion metrics, including accuracy, precision, and the F1 score. In addition, we have compared the performance of the proposed model with the models proposed in the previous study, which were built based on the received signal strength (RSS). The results show that the effectiveness of the models improves as the number of base stations increases.

Deep neural network (DNN)-based vehicle detection systems deployed on unmanned aerial vehicles (UAVs) are susceptible to adversarial attacks, resulting in significant implications for public safety and system reliability. Despite advancements in DNN-based detection, the adversarial robustness of these systems in aerial video contexts remains underexplored. Existing attack models fail to exploit the sequential and periodic nature of video frames in aerial vehicle detection systems. To address this, we propose a Periodic Adversarial Attack for Aerial Video (P3AV), which is the first to take advantage of the periodic nature of tasks related to road traffic parameters and improve the success of attacks. P3AV systematically selects critical video frames to be attacked by employing Bayesian optimization combined with domain-specific knowledge. The sensitive pixels in the frames are then chosen based on the gradient magnitudes of the loss function. Finally, an improved version of the projected gradient descent algorithm is developed by using gradient norms to generate perturbations and enhance the manipulation of selected pixels. Our experiments using four adversarial attacks against 10 DNN architectures, which are developed based on Convolutional Neural Network (CNN) and YOLO, on two datasets demonstrate that P3AV can improve the false rate in detection systems by 6% and the attack success rate by 5% over other attack models. Meanwhile, CNN models perform the worst against adversarial attacks. These findings highlight the critical need for improved adversarial defenses in UAV-based detection systems and underscore the broader implications for secure and reliable ITS.

• We propose MVGFormer, a multi-view fusion framework with a dual-stream encoder guided by a visual graph. • We design two decoder variants: MF for hierarchical feature fusion and P3DA for multi-scale representation. • We introduce a view-masked self-supervised strategy that reconstructs masked views from remaining ones. • MVGFormer achieves superior performance over existing state-of-the-art cryo-ET segmentation methods. Cryo-Electron Tomography (cryo-ET) is a cutting-edge 3D imaging technology that enables detailed examination of biological macromolecular structures at near-atomic resolution. Recent deep learning applications on cryo-ET, such as cryo-ET segmentation, have drawn widespread interest for their potential to improve particle alignment, classification, and other tasks. However, current methods heavily rely on convolutional architectures, which prioritize local information while neglecting the global structural information inherent in cryo-ET data. Transformer-based models, known for their large receptive field, have become the de-facto design for 2D vision tasks due to their ability to effectively capture global information. This approach is also well-suited for 3D tasks, given the complex nature of 3D objects. Based on this, we extend 2D vision transformers into 3D and propose a novel transformer-based framework for cryo-ET segmentation, named MVGFormer. MVGFormer introduces a multi-view perspective fusion transformer encoder, which captures rich global structural information from multiple perspectives using unique positional embeddings. To enhance contextual awareness, we design a parallel context encoder that builds a visual graph to guide attention. We further introduce two complementary 3D decoders: multi-level feature fusion (MF) and parallel atrous convolutions (P3DA), which together capture multi-scale structural cues for precise segmentation. Furthermore, we introduce a view-masked self-supervised learning strategy to reinforce the effectiveness of the multi-view design and improve the model’s representation capability. To our knowledge, MVGFormer is the first transformer-based model for cryo-ET segmentation. We empirically evaluate MVGFormer on six cryo-ET datasets across three different tasks. Extensive experimental results demonstrate its superiority over state-of-the-art 3D segmentation methods.

Urban Rail Transit (URT) accidents not only seriously affect the safety and reliability of its operations, but also reduce service level to passengers. Based on historical URT accident data, this study develops a hybrid ensemble learning model based on a Convolutional Neural Network (CNN) and Bidirectional Encoder Representations from Transformers (BERT) for predicting accident consequences in URT. The CNN is employed to capture spatial patterns from the diverse accident data, while the BERT is applied to learn complex relations in accident text descriptions. The results of the two models are combined for classifying accident consequences. The proposed hybrid ensemble learning model was applied to predict accident consequences in Chongqing’s URT using historical accident records. It achieved a prediction accuracy of 0.805 on testing data set, which is at least 20% higher than that of commonly used machine learning models, including multilayer perceptrons, support vector machines, and Bayesian networks. Furthermore, the reapplication of the proposed model to historical accident records of the URT in Chengdu demonstrates the generalizability and reusability of the model. This study forecasts the consequences of URT accidents with high accuracy using limited historical data, which supports operators in identifying high-frequency and high-impact accidents. Consequently, targeted maintenance and timely emergency response strategies can be developed to decrease accident rates and mitigate the impacts.

Dynamic ridesharing has gained significant attention in recent years. However, existing ridesharing studies often focus on optimizing order dispatching and vehicle repositioning separately, leading to short-sighted decisions and underutilization of the ridesharing potential. In this paper, we propose a novel joint optimization framework called$\mathtt {JODR}$. By coordinating order dispatching and vehicle repositioning,$\mathtt {JODR}$enhances ridesharing efficiency while ensuring high-quality service. The core idea of$\mathtt {JODR}$is to dispatch ride orders with high demand in specific mobility directions to vehicles with sufficient available capacity, effectively balancing future supply and demand in those directions. To achieve this, we introduce a novel mobility value function that can predict the long-term mobility value of matching an order with its travel direction. By considering orders’ directional mobility values, service quality assessments, and available vehicle capacities,$\mathtt {JODR}$formulates the order dispatching as a minimum-cost maximum-flow problem to derive the optimal order-vehicle assignments. Furthermore, the value function helps the intelligent repositioning of idle vehicles. Extensive experiments conducted on a large real-world dataset demonstrate the superiority of$\mathtt {JODR}$over state-of-the-art methods across various performance metrics. These experimental results validate the effectiveness of$\mathtt {JODR}$in improving the ridesharing efficiency and experience.

Unmanned Aerial Vehicles (UAVs), or drones, are increasingly used in various fields. A major concern with UAV operation is their limited power capacity which impacts mission planning, operational efficiency, and battery management, presenting significant research and engineering challenges. This paper evaluates the applications of multiple AI algorithms in predicting the energy consumption of low-cost quadcopter drones. One of the primary contributions involves developing four prediction models, including random forest, regression tree, support vector machine, artificial neural network, and adaptive Neuro-Fuzzy Inference System (ANFIS) on an open-source dataset of small quadcopter flights. This paper also performs a comparative study on the performance of the aforementioned algorithms in predicting the energy consumption of a UAV. This research enhances the field not only by leveraging established machine learning techniques but also by adopting and examining ANFIS, which has received limited prior research attention. By introducing and applying ANFIS, this study not only expands the existing knowledge but also offers a unique perspective, potentially paving the way for further research, especially in addressing uncertainty like weather conditions. According to our study, the power consumption of the UAV is notably influenced by the aircraft’s altitude, wind speed, and velocity. The Random Forest model demonstrates superior accuracy in forecasting UAV power consumption compared to other models. We also provide an overview of the ongoing challenges and potential future endeavors.

The increased reliance on Unmanned Aerial Vehicles (UAVs) in various industries exalts the security requirements since it is critical to protect these systems from any cyber-attack. GPS spoofing presents an important challenge by deceiving UAVs through false GPS signals that would disrupt their operations, thereby endangering them. As a countermeasure, this study introduces a method of detecting GPS spoofing attacks that are aimed at UAV systems. This involves developing a robust methodology to detect the GPS spoofing attack based on the UAV’s current reported location and Received Signal Strength (RSS) data at several base stations. In this study, we developed a smart verification algorithm using the K-Nearest Neighbors (KNN) algorithm to authenticate the reported locations of UAVs, based on RSS from various base stations antenna. We evaluated the performance of the algorithm using metrics such as accuracy, precision, and F1-score. The results indicate that the algorithm’s effectiveness improves with an increase in the number of base stations used. Additionally, the paper will pinpoint the possible direction for UAV security and the adaptive countermeasures to improve the level of resilience against spoofing tactics, which are rapidly evolving.

Disruptions occurring at an urban rail transit (URT) system can severely affect its normal operations, and an effective bus bridging service (BBS) is able to help to reduce the negative effects. Transit operators usually arrange BBS to depart from the disrupted stations to evacuate the stranded passengers. However, the overload of passengers at the disrupted stations, especially at turnover and transfer stations, may incur the secondary operation disruptions such as stampede accidents. To mitigate the negative effects of disruptions and reduce the number of passengers stranded at the disrupted stations, the strategy based on the passenger guidance and extended BBS (E-BBS) is introduced in this paper. Different from the widely applied standard BBS running within the disrupted links, E-BBS runs among the normal operating stations or between the normal operating stations and the disrupted stations. A mixed integer linear programming model is developed in this paper to guide the stranded passengers and design an optimal E-BBS solution to transport them. Considering the bi-directional running trains along the disrupted links, a dynamic decision framework is developed to manage the disruptions. Given the impacts of the high uncertainty on bus running time which could affect the performance of E-BBS, a robust model is proposed to obtain more reliable travel guidance and BBS schemes. Numerical experiments based on Chengdu subway in China are conducted. The results indicate that the proposed model can obtain optimal travel guidance and E-BBS solutions in a timely manner. When the uncertainty on bus running time is ignored, the total travel cost for affected passengers is reduced 14.20 % on average with the aid of the optimal travel guidance and E-BBS solutions. Moreover, the number of passengers gathering at the disrupted stations decreases by 36.80 %. The robust model can obtain more reliable travel guidance and E-BBS schemes in consideration of uncertain bus running time. The proposed model shows great potential to effectively mitigate the negative effects of disruptions and help to enhance the capability of a URT system to respond to disruptions.

Unmanned Aerial Vehicles (UAVs) have immense potential to enhance Intelligent Transport Systems (ITS) by aiding in real-time traffic monitoring, emergency response, and infrastructure inspection, leading to rich data collection, lower response times, and efficient urban mobility management. Machine learning (ML) is a crucial component in UAV-assisted ITS as it processes UAV-captured data in both the perception layer and decision layers of intelligent components for vehicle/pedestrian detection, trajectory optimization, and resource allocation. Importantly, the integration of UAVs and cutting-edge deep learning (DL) techniques is fostering an exciting synergy, equipping UAVs with unparalleled intelligence and autonomy, particularly, for the perception layer of UAVs. Despite these enhancements, their usefulness for detection and traffic extraction tasks remains largely unexplored. The contributions of this paper are divided into two main aspects: (1) UAVs in different ITS application scenarios that are empowered by ML technologies are reviewed. (2) A thorough survey aiming to explore a quantitative understanding of widely used DL models via a series of experiments and comparisons is presented. Four DL models, namely Convolution Neural Network (CNN), regions with CNN (R-CNN), Faster R-CNN, and You Only Look Once (YOLO)), in combination with different backbones, are designed and employed on five aerial datasets. Finally, we present a discussion of the remaining challenges and future works.

In the public transportation system, punctuality benefits both bus operation and passengers’ travel experience. However, uncertainty exists due to complex traffic conditions and heterogeneous driving behaviors. To analyze bus operational uncertainty, transport planners and bus operators need a tool that supports multi-granular modeling, spatio-temporal representation, and interactive exploration. To meet the requirement, we present MetroBUX, a visual analytics system for$B$us operational$U$ncertainty e$X$ploration. MetroBUX aligns daily bus trips and models stop-level uncertainty of bus arrival time. It has a consolidated interface with three main views: Map View for presenting the spatial distribution of uncertainty, Temporal View for tracking the evolution of uncertainty, and Trip View for inspecting uncertainty propagation. Specifically, MetroBUX enables integrated spatio-temporal analysis by connecting topological uncertainty distribution at different periods in a nested tracking graph. Furthermore, it supports interactive and hierarchical exploration, including region-, route-, trip-, and stop-level analysis. Case studies on real-world bus operational data and domain experts’ feedback demonstrate the efficiency of MetroBUX.

Traffic signal control (TSC) is still one of the most significant and challenging research problems in the transportation field. Reinforcement learning (RL) has achieved great success in TSC but suffers from critically high learning costs in practical applications due to the excessive trial-and-error learning process. Offline RL is a promising method to reduce learning costs whereas the data distribution shift issue is still up in the air. To this end, in this paper, we formulate TSC as a sequence modeling problem with a sequence of Markov decision process described by states, actions, and rewards from the traffic environment. A novel framework, namely TransformerLight, is introduced, which does not aim to fit into value functions by averaging all possible returns, but produces the best possible actions using a gated Transformer. Additionally, the learning process of TransformerLight is much more stable by replacing the residual connections with gated transformer blocks due to a dynamic system perspective. Through numerical experiments on offline datasets, we demonstrate that the TransformerLight model: (1) can build a high-performance adaptive TSC model without dynamic programming; (2) achieves a new state-of-the-art compared to most published offline RL methods so far; and (3) shows a more stable learning process than offline RL and recent Transformer-based methods. The relevant dataset and code are available at Github.

An optimal train schedule enables the railway system to serve maximal passengers with limited resources. However, inappropriate train services may cause a diversity of adverse impacts such as service delay and long waiting times. This study aims to improve train service with a new schedule compared to the existing train schedule. To develop an alternative train service with a schedule solution, a mixed integer linear programming model is developed. A comparison between the existing train service and an alternative service is conducted based on a subset railway network in New South Wales, Australia. Numerical results proved that the alternative train service outperforms the existing one with regard to delay time and train capacity.

Many studies confirmed that a proper traffic state representation is more important than complex algorithms for the classical traffic signal control (TSC) problem. In this paper, we (1) present a novel, flexible and efficient method, namely advanced max pressure (Advanced-MP), taking both running and queuing vehicles into consideration to decide whether to change current signal phase; (2) inventively design the traffic movement representation with the efficient pressure and effective running vehicles from Advanced-MP, namely advanced traffic state (ATS); and (3) develop a reinforcement learning (RL) based algorithm template, called Advanced-XLight, by combining ATS with the latest RL approaches, and generate two RL algorithms, namely "Advanced-MPLight" and "Advanced-CoLight" from Advanced-XLight. Comprehensive experiments on multiple real-world datasets show that: (1) the Advanced-MP outperforms baseline methods, and it is also efficient and reliable for deployment; and (2) Advanced-MPLight and Advanced-CoLight can achieve the state-of-the-art.

Operational disturbances within 30 min, namely, short-term operational disturbances (STODs), occur frequently during the daily operation of an urban rail transit (URT) system. Therefore, there is an urgent need to assess a network’s ability to respond to STODs to improve its operational level. The resilience of a URT network jointly considering turn-back operations, the occurrence time of STODs, and passenger travel experience is addressed. By considering passenger travel alternatives under STODs, we assessed the resilience of a network by utilizing the ratio of the average loss of a time-dependent performance indicator, which is referred to as the operational service level indicator in this paper. A simulation-based resilience assessment flowchart is also proposed. Numerical experiments conducted on the Chengdu subway network indicate that this network is more resilient to successive station failure than to simultaneous failure. Guaranteeing the normal operation of transfer stations is vital for enhancing the resilience of a network. The metric proposed in this paper is more practical for assessing resilience than the two commonly used metrics (performance loss and disturbance duration). Additionally, the experimental results highlight that turn-back operations at the stations have a significant impact on the resilience of a URT network. The sensitivity analysis of the parameters indicates that ensuring the normal operation of links connected by critical stations, preventing large-scale station failure, improving passenger tolerance to disturbance durations, and restoring the connectivity of disrupted links rapidly have practical significance for enhancing the resilience of a network.

As a common phenomenon, overtaking behaviour is frequently observed on pedestrian flow, which not only reshapes pedestrian flow but also generates adverse impacts on pedestrian safety to some extent. Prior research focused on unidirectional pedestrian modelling, especially with overtaking behaviour, is limited. Moreover, pedestrian behaviour in the context of COVID-19 is rarely investigated. Inspired by the social force model, this paper proposes a dynamic sensitivity model for unidirectional pedestrian flow, which is able to describe the overtaking behaviour and analyse the potential impact of COVID-19 on pedestrian behaviour. In the proposed model, dynamic sensitivity and attention field of pedestrians are introduced to embody the effects of individual characteristics and surrounding environments on pedestrian behaviours. To calibrate the model and evaluate the effects of COVID-19 pandemic on pedestrian dynamics, real-life data collected by video recordings in Nanjing, China is used in this study. The simulation results indicate that the dynamic sensitivity model is able to reflect the variance of the adaptive velocity and route choice of overtaking pedestrians on unidirectional pedestrian flow. Our research findings show that the social distance during COVID-19 is higher than the value under normal conditions, and the majority of pedestrians tend to follow the suggested social distancing rules during COVID-19. Moreover, the overtaking pedestrians violate the suggested social distancing rules more frequently than the rest pedestrians.

Different from the existing train delay studies that had strived to explore sophisticated algorithms, this paper focuses on finding the bound of improvements on predicting multi-scenario train delays with different machine learning methods. Motivated by the observation of deep learning methods failing to improve the prediction performance if the delay occurs rarely, we present a novel augmented machine learning approach to improve the overall prediction accuracy further. Our solution proposes a rule-driven automation (RDA) method, including a delay status labeling (DSL) algorithm, and the resilience of section (RSE) and resilience of station (RST) indicators to generate the forecast for train delays. The experiment results demonstrate that the Random Forest based implementation of our RDA method (RF-RDA) can significantly improve the generalization ability of multivariate multi-step forecast models for multi-scenario train delay prediction. The proposed solution surpasses state-of-art baselines based on real-world traffic datasets, which treat various real-time delays differently. Even when the predictability of conventional deep learning methods decreases, the performance of our method is still acceptable for practical use to provide accurate forecasts.

Bus bunching, a phenomenon due to the failure of headway or timetable adherence, often causes low level of public transit service with poor bus on-time performance and excessive passenger waiting time. To mitigate bus bunching, an accurate and real-time prediction method plays an important role. In this paper, we propose a supply-demand seq2seq model called SD-seq2seq to predict bus bunching using smart card data. Features from both supply and demand sides of bus service are taken into account, like bus stop type, dwelling time, passenger demand and type, and so on. Extensive experiments on multiple bus routes in real world demonstrate that our method outperforms other baseline methods. The proposed method is expected to provide useful online information of bus operation to both bus operators and passengers.

Bus bunching is a result of sophisticated traffic condition, unstable bus operation and dynamic travel demand, which not only causes passengers' dissatisfaction, but also degrades the bus service performance. To tackle the bus bunching issue, multiple steps are usually adopted, from bus bunching identification to solution development. This paper serves as the first-mile work of a series of studies, to identify bus bunching in a large-scale public transport network. Selected findings from a case study of the Greater Sydney Region based on smart card data analysis are shown to illustrate the bus bunching issues at multiple levels.