Engineering

Abstract: Bridge management systems primarily prioritize structural condition, while pavement management systems focus on roadway performance. Neither provides a unified measure of the potential for bridges to impede traffic movement by jointly considering infrastructure condition and traffic demand. This separation limits the ability of stakeholders to identify bridges where preservation investments may yield the greatest mobility benefits across a national transportation network. This study develops a mobility impedance index (MII) by integrating pavement roughness, bridge deck condition, current traffic demand, and projected traffic growth into a single composite measure. A transparent and transferrable data-integration and prioritization methodology was developed to spatially integrate the Highway Performance Monitoring System (HPMS) and National Bridge Inventory (NBI), producing a national analytical dataset of 79,595 open, unrestricted mainline NHS bridges in the contiguous United States. The index was evaluated using distributional analysis, local Getis-Ord hotspot analysis, sensitivity testing, and temporal comparability assessment. The MII exhibited a well-behaved generalized normal distribution and identified 20,621 bridges as statistically significant local hotspots, concentrated primarily within major metropolitan regions and high-demand transportation corridors. Sensitivity analyses demonstrated that the proposed MII is robust, with bridge-rank correlations ranging from 0.807 to 0.965 across the alternative specification scenarios and hotspot-set similarities ranging from 0.446 to 0.801. Restricting the analysis to bridges whose component observation years were within three years produced virtually identical prioritization results (Spearman ρ = 0.999), confirming that temporal differences among the component datasets have negligible influence on bridge rankings. The proposed MII provides transportation agencies with a nationally scalable data-driven decision-support method for network-level bridge prioritization that complements existing bridge and pavement management practices by identifying bridges and regions where preservation investments are likely to provide the greatest mobility benefits and support more sustainable allocation of limited infrastructure preservation resources.

Abstract: Transportation network disruptions can abruptly alter travel costs and trigger day-to-day route-choice adjustments. Conventional day-to-day dynamic traffic assignment models, however, often rely on fixed-weight or uniformly decaying memory structures, limiting their ability to represent traveller learning under non-stationary conditions. This study develops a Bayes with Adaptive Memory (BAM)-based framework that selectively retains, down-weights, or discounts past travel-cost experiences according to recency and statistical significance. The BAM mechanism is integrated into a link-based day-to-day traffic assignment model to examine how adaptive memory influences perceived costs, route-choice adjustment, and post-disruption traffic-flow evolution. Numerical experiments on a hypothetical network show smoother flow trajectories, faster stabilisation, and a 41.2% reduction in mean absolute perception-tracking error compared with a conventional fixed-weight model. Under a link-removal scenario, the framework captures immediate traffic redistribution followed by gradual stabilisation toward a new post-disruption state. Sensitivity analysis reveals a trade-off between responsiveness and stability: shorter memory improves cost tracking, whereas longer memory and stronger significance weighting promote smoother adjustment and earlier convergence. Within the tested setting, the results suggest that adaptive memory provides a behaviourally plausible mechanism for representing traveller learning and may support disruption-management assessments of adjustment periods, congestion redistribution, and information provision.

Abstract: Millimeter-wave radar is an attractive sensing modality for autonomous driving because of its robustness under adverse weather and lighting conditions. However, comparisons between radar object detectors are hindered by inconsistent evaluation procedures and limited reporting of deployment costs. This work presents a comprehensive benchmark using standardized VOC mean average precision (mAP) and evaluates end-to-end latency across preprocessing, network inference, and postprocessing. We evaluate SSRaDNet, a lightweight hybrid CNN–Swin Transformer detector, across analog-to-digital converter (ADC), range-Doppler (RD), and range-azimuth-Doppler (RAD) representations and compare it with existing methods on RADIal and RADDet. SSRaDNet achieves state-of-the-art AP@0.5 among re-evaluated RADIal models using RD input, reaching 88.89%, and provides the strongest overall balance of detection and segmentation performance among the compared ADC configurations. On RADDet, it achieves the best performance among methods using the native-resolution RAD tensor (45.69% mAP@0.5) and introduces the first ADC-based detector with full bounding-box regression. We also evaluate alternative postprocessing strategies, including hybrid MaxPoolNMS-GreedyNMS, to characterize deployment trade-offs on resource-constrained hardware. These results demonstrate the importance of standardized evaluation and full-pipeline runtime analysis for radar-based object detection. Code is available at https://github.com/IqbalBan/SSRadNet.

Abstract: Driving assistance and Autonomous driving are among the fastest evolving domains of intelligent transportation systems (ITSs). However, visually impaired pedestrians (VIPs), remain weakly protected by the roadsides or vehicle perception systems, especially when the right of way must be communicated in an accessible and auditable manner. In this paper we describe a safety-constrained vehicle-to-pedestrian (V2P) architecture designed for VIPs crossing assistance. The system creates a time bounded interaction state between the pedestrians, the roadside infrastructure, and the vehicles at Fog, to avoid permissive instructions before all the safety stacks are respected in each level where there are formal time bounded strictures with appropriate tools to formally abide by these rules during each interaction between vehicles and VIPs. Our objective is not to design a new detector, tracker, or MLLM model. Instead, we combine edge descriptors, fog level multi-view descriptors, and multi-camera multiple objects tracking technique (MC-MOT) within a latency-constrained V2P loop. Our main empirical focus is the WildTrack multi-camera continuity, with a novel Multi-Camera Consensus Gating (MCCG) safety primitive criterion; we also conducted a JAAD experiment as secondary assistive-branch for feasibility check, indicating plausible crossing-intent estimates, sub-second restrictive alerts, and actionable grounded guidance. On WildTrack the reproduced Fog continuity branch reaches a MOTA = 88.8%, IDF1 = 91.7%, and HOTA = 65.5%. In the WildTrack settings with N = 7, MCCG with M = 4 blocks 75% of ID-switch considering a 5.0% permissive-eligibility cost. These findings support feasibility of the proposed ITS interaction loop, while leaving field deployment of V2X stack validation, and user studies for future work.

Abstract: Car-sharing systems, as an important form of shared transport, face supply-demand imbalance and complex scheduling challenges. Heterogeneous vehicle can better meet diverse customer demands by offering vehicles with different capacities. Meanwhile, uncertainty in dispatcher movement time and vehicle relocation time may affect the reliability of scheduling plans. To address these issues, this paper focuses on a one-way station-based car-sharing problem, considering integrated scheduling, heterogeneous vehicle, uncertain dispatcher movement time and vehicle relocation time simultaneously. We define the uncertainty sets for dispatcher movement time and vehicle relocation time based on historical data from the Taxi & Limousine Commission (TLC) in Queens, New York by data-driven method. A robust optimization model is developed to improve order fulfillment and scheduling stability. We develop a hybrid approach that combines adaptive large neighborhood search and a genetic algorithm (ALNS/GA), and incorporates robustness into solution decoding and fitness evaluation under the data-driven uncertainty sets. Numerical experiments based on real TLC data show that the integrated scheduling system reduces average costs by 10.69\% compared with non-integrated systems, while the integrated robust scheduling system further improves the order fulfillment rate by 0.99\%. Sensitivity analysis also offers useful managerial insights.

Abstract: Electric vehicles (EVs) are increasingly recognised as a key component of sustainable mobility transitions. While technological advancements and supportive policies have significantly expanded EV adoption, the relatively high purchase cost of conventional EVs remains a barrier for many consumers. In response, a new generation of frugal EVs has emerged, emphasizing affordability, resource efficiency, and functional adequacy while maintaining essential mobility requirements. Despite growing commercial interest, the factors influencing the acceptance and affordability of frugal EVs remain underexplored in the academic literature. This paper addresses this gap through a structured literature review using the SALSA (Search, Appraisal, Synthesis, and Analysis) framework. A total of 116 studies were selected and analysed to identify the key factors influencing user acceptance and affordability in the context of frugal EVs. The reviewed evidence was synthesised into four principal dimensions: demographic, situational, contextual, and psychological factors. The findings indicate that affordability, low operating costs, compact urban mobility, modular vehicle design, and battery-swapping technologies are among the most important drivers of frugal EV adoption. In contrast, safety concerns, limited driving range, charging infrastructure constraints, weak resale value, regulatory uncertainty, and perceptions of lower quality remain significant barriers to market acceptance. The review further suggests that the adoption dynamics of frugal EVs differ from those of conventional EVs, particularly due to their emphasis on cost-performance trade-offs and urban mobility applications. From a smart-city perspective, frugal EVs have the potential to support more inclusive and sustainable urban transportation systems by expanding access to low-carbon mobility for cost-sensitive users and underserved market segments. The study contributes to the emerging literature on affordable electric mobility by providing a dedicated framework for understanding frugal EV acceptance and affordability. It concludes with recommendations for policymakers, manufacturers, infrastructure providers, and service operators aimed at supporting the wider adoption and distribution of frugal EVs.

Abstract: As the development of deep oil and gas resources advances continuously, the role of weighted fracturing fluids in reducing wellhead operating pressure becomes increasingly important. However, the turbulent frictional drag behaviors of weighted fracturing fluids in pipes have not been understood systematically. The frictional drag characteristics and flow behaviors of non-weighted and weighted fracturing fluids in pipes are investigated at different flow rates, densities, pipe diameters, and rheological parameters via steady-state rheological tests and pipe flow friction experiments, together with large-eddy simulation and a power-law constitutive model for generalized Newtonian fluids. Main findings are as follows. The drag reduction rates of non-weighted polyacrylamide solutions basically increase with flow rate, and the optimal molecular weight of 600–700×10⁴ corresponds to the highest drag reduction rate of 62%, while excessively large molecular weights weaken drag reduction effectiveness due to enhanced chain entanglement. The types of weighting salts have a significant effect on drag reduction performance. The 20% KCl system obtains the highest drag reduction rate of 64.8%, while the high-density CaBr₂ system gets the lowest of 42.3% because high salinity suppresses the stretching of polymers. Parameter sensitivity analysis indicates that the power-law index n is the most sensitive parameter controlling frictional drag, followed by pipe diameter D. The novelty of this study lies in integrating experimental investigation, LES, and parameter sensitivity analysis to provide a theoretical basis for deep fracturing working conditions and fracturing fluids design.

Abstract: Automatic incident detection (AID) can significantly decrease the time to detect incidents and better identify incident locations to allow faster response. The deployment of traffic surveillance cameras has motivated transportation agencies to investigate computer vi-sion technologies for AID using live video streams from these cameras. This paper evaluated three commercially available computer vision-based AID systems. The evaluated AID systems were able to use video streams from six existing Pan-Tilt-Zoom (PTZ) surveillance cameras for incident detection when these cameras were on predefined presets. The three systems with the selected settings performed differently. For example, the system with the highest number of true stopped vehicle incident detections also had a relatively high number of false alerts. The AID system accuracy also varied by study site, likely due to traffic patterns, roadway geometry, and camera angles. Two major factors contributed to false alerts: stop-and-go traffic during non-incident conditions and other highway objects, including light reflection and glare, that were falsely detected as incidents. The three systems also detected additional incidents not reported in existing incident management databases, but these were mainly shoulder incidents and work zone vehicles.

Abstract: Reliable short-term position forecasting can support collision-risk assessment, communication continuity, and prediction-assisted control in Internet of Things (IoT)-enabled unmanned aerial vehicles (UAVs). This study reformulates UAV position prediction as a flight-wise, multi-horizon, three-dimensional forecasting problem and tests whether position, velocity, gravity-resolved acceleration, and quaternion-orientation histories improve predictive accuracy while preserving edge feasibility. The dataset contains 3100 AirSim flights with high-rate kinematic, inertial, attitude, pressure, and magnetic-field measurements under variable horizontal wind. Signals are converted to a common navigation frame, gravity-resolved, low-pass filtered, resampled to 50 Hz, and partitioned by flight identifier before normalization and window construction. Each learned model receives 2 s of history and predicts the complete next 1 s trajectory, with errors evaluated at 0.1, 0.5, and 1.0 s. The sensor-enriched LSTM (LSTM-PVAQ) is compared under matched conditions with persistence, constant-velocity, constant-acceleration, extended Kalman filter, reduced-feature LSTM, GRU, temporal convolutional network (TCN), and compact Transformer baselines. LSTM-PVAQ achieved 3D RMSE values of 0.043, 0.168, and 0.371 m at 0.1, 0.5, and 1.0 s, respectively. At 1 s, its RMSE was 21.7% lower than LSTM-PV, 13.1% lower than GRU-PVAQ, 9.3% lower than TCN-PVAQ, and 16.8% lower than Transformer-PVAQ. Its one-second ADE and FDE were 0.216 and 0.339 m. On a Raspberry Pi 5 CPU using one FP32 thread and batch size one, median inference latency was 0.88 ms, well below the 20 ms model-update interval. The results show that gravity-resolved inertial and orientation histories improve multi-horizon prediction, while TCN-PVAQ remains an attractive lower-latency alternative.

Abstract: The traditional truck-based urban logistics drives the highest unit costs and externalities in urban logistics, including elevated CO₂ emissions, air pollution, road accidents, traffic congestion, road damage, and adverse social effects on delivery workers and communities. In Italy, the unique urban morphology of historic city centers, coupled with stringent Limited Traffic Zones (ZTL), make the impacts particularly severe. This paper proposes a hybrid system that exploits the existing electric regional rail network to transport freight to urban areas with fewer externalities. The system utilizes autonomous, electric wagons, such as the type built by Parallel Systems, a start-up in Los Angeles, that couple with passenger trains for suburban-to-urban transit and decouple to travel independently over dedicated short-distance connectors to urban cross-docking terminals. The study aims to improve delivery efficiency while reducing the environmental impact. A preliminary analysis for the city of Rome shows the possibility of implementing the system, and the findings indicate that Autonomous Electric Wagons (AEW) have the potential to provide a scalable and environmentally sustainable alternative for urban logistics.

Abstract: Evaluating multiple-object trackers is challenging due to the variable number of quantities involved and the mixed discrete--continuum nature of the problem. Existing methodologies primarily address detection and tracking challenges. These challenges aim at whole computer-vision pipelines as opposed to isolated tracker algorithms. However, modern tracker algorithms have become sufficiently sophisticated to merit a stand-alone analysis. The most critical component of a tracker is the association procedure, as its outcome strongly affects the overall tracking quality. We propose a straightforward quality assessment framework to evaluate the association quality of trackers. The framework relies on a minimal, query-oriented instrumentation of the tracker. This instrumentation exposes the tracker's internal association decisions, allowing for a binary classification of detection-target matches. The proposed methodology is simple to implement, introduces minimal computational overhead, and agrees with the well-known ClearMOT metrics. The comparison is implemented through our accompanying open-source software packages.

Abstract: This study presents a data-driven Digital Twin framework designed to optimize closed-loop Adaptive Cruise Control (ACC) operational parameters for eco-driving without modifying electronic control unit (ECU) firmware. High-frequency telematics data (OBD-II and GPS) were recorded from a production vehicle over an 11.57 km test corridor to engineer and calibrate a virtual Digital Twin in PTV Vissim using Wiedemann 99 car-following logic. The validated model demonstrated high empirical fidelity, achieving a mean velocity error of 2.84 km/h, a GEH statistic of 2.67, and an absolute fuel consumption variance of just 0.01 L/100 km. An iterative parametric optimization (~100 simulation runs) was executed to tune acceleration limits, convergence time constants, and target velocity bounds while maintaining strict safety constraints. The optimized Eco configuration achieved a 12% reduction in fuel consumption (4.91 to 4.33 L/100 km) alongside a 14% increase in trip duration (893 to 1019 s). Bivariate state-space analysis confirmed the complete elimination of high-intensity accelerations and a 0.48 percentage point reduction in hard braking. The framework serves as an algorithmic decision support tool for energy-aware fleet management and adaptive vehicle control.

Abstract: In recent years, increasing attention has been devoted to improving cyclists’ safety in road traffic, particularly on roads without separated cycling infrastructure. Advisory cycle lanes (2–1 roads) have recently been introduced in several European countries on narrow roads where conventional cycling facilities cannot be provided. However, evidence regarding their safety effects remains limited. This paper evaluates the effects of advisory cycle lanes on overtaking behaviour and crash occurrence on selected roads in and around Maribor, Slovenia. The study combines OpenBikeSensor measurements of lateral passing distance, traffic volume and speed data, and a before-and-after analysis of police-reported crashes. In total, 970 overtaking manoeuvres were recorded on six streets with advisory lanes and comparable control streets without cycling infrastructure. Advisory bike lanes were associated with greater lateral passing distances, with mean clearance increasing from 135 cm to 156 cm. Nevertheless, around 40–50% of overtakes still occurred within 1.5 m of the cyclist. Crash analysis indicated reductions of about 22% in total crashes and 38% in cyclist-involved crashes, although the small number of crashes warrants cautious interpretation. The collected data enable a comparison of traffic behaviour and safety effects of advisory cycle lanes in Slovenia, providing a basis for future planning of traffic infrastructure aimed at improving the coexistence of cyclists and motor vehicle drivers.

Abstract: Road traffic accident remains a major public-health and transportation challenge in Bangladesh, while existing studies often rely on isolated datasets, random validation, point predictions, and descriptive rankings that do not distinguish absolute burden from exposure-adjusted risk. This study develops an integrated, leakage-safe, uncertainty-aware, and policy-oriented framework using four nationwide datasets covering population characteristics, accidents, fatalities, injuries, vehicle involvement, and vehicle-specific fatalities across eight administrative divisions from January 2023 to December 2025. Linked division–month and vehicle–division–month panels were constructed with population-normalized indicators, cyclical seasonality, lagged dynamics, rolling variability, and momentum features. Mean and seasonal-naïve baselines, Ridge, Poisson, Tweedie, Random Forest, Extra Trees, Gradient Boosting, Histogram Gradient Boosting, and a hurdle model were evaluated using expanding-window validation and an independent 2025 holdout year. The framework further incorporated empirical-Bayes vehicle-risk stabilization, split-conformal prediction intervals, permutation importance, feature-family ablation, spatial concentration, hotspot persistence, and a transparent policy-priority index. National fatalities remained persistently high, while exposure-adjusted analysis identified Barishal and Sylhet as emerging-risk divisions despite the larger absolute burden in Dhaka and Chattogram. Motorcycles recorded the highest stabilized fatality risk, followed by auto rickshaws and other lightly protected vehicle categories. Random Forest achieved the lowest holdout fatality RMSE of 15.77, outperforming the seasonal-naïve benchmark by 26.7%, although count and regularized models remained competitive. Ablation analysis showed that the temporal-plus-demographic specification generalized better than the fully integrated feature set. The study contributes a reproducible decision-support architecture for seasonal enforcement, vehicle regulation, emergency-response allocation, infrastructure prioritization, and divisional monitoring. Prospective validation remains necessary before full operational deployment.

Abstract: Inter-hub feeder service is a promising application for electric vertical takeoff and landing aircraft (eVTOL), especially when passengers must connect to subsequent flights under tight time constraints. This study develops an optimization framework that accounts for heterogeneous passenger behavior. Based on a stated-preference survey, we introduce delay-risk perception under remaining connection time constraints into passenger utility, identify heterogeneous preference classes, and formulate a bilevel optimization model. The upper level selects eVTOL schedules under given resource and fare configurations, while the lower level captures the stochastic user equilibrium of heterogeneous passengers competing across eVTOL and external transport alternatives. To solve the resulting mixed-integer nonlinear bilevel problem, we propose a Neural Bilevel Optimization and generalized Benders decomposition (Neur2BiLO-GBD) hybrid algorithm. Numerical experiments for the Shanghai Hongqiao-Pudong inter-hub transfer corridor show that the profit-maximizing operating plan generates positive social net utility of the feeder system under the baseline setting. Fleet size, charging infrastructure, and fare levels affect operator profit and social net utility differently, and their high-value regions do not fully coincide across operating scenarios. When external transport faces larger potential delays and remaining connection time is short, eVTOL is more likely to achieve both high operator profit and high social net utility.

Abstract: Urban road networks in sub-Saharan Africa diverge from traditional sensor-rich, lane-disciplined environments. This paper established an empirical baseline assessment of network efficiency and velocity decay along the 8.0km Mile 17 to Governor's Roundabout corridor (8.0 km) in Buea, Cameroon. Characterized by radial geometry, a steep 25 m/km inbound gradient, unsignalized intersections, and a high mix informal transport (taxis comprising 59.5% and 65.27 of traffic for inbound and outbound flow, respectively). Fifteen-minute Passenger Car Unit (PCU) volume counts were recorded at three measurement points evaluated across three peak windows (morning, afternoon, and evening) over three consecutive weekday observation days. Volume-to-capacity ratios (v/c), peak-hour factors (PHF), equivalent hourly flow rates, space-mean speeds, and Mean Travel Time Indices (MTTI) were calculated for both inbound and outbound flow directions. Results reveal near-saturation and over-saturation at the Bonduma mid-corridor point during morning and afternoon peaks (v/c: 0.98 to 1.06). The highest operational stress occurred during the afternoon outbound window, where over-capacity conditions emerged simultaneously at Mile 17 and Bonduma. MTTI values ranged from 1.36 to 2.04, indicating that congestion is primarily driven by informal transport dynamics rather than absolute vehicle volume. Finally, the paper introduces the Directional Flow Asymmetry Index (DFAI) and Corridor Velocity Decay Rate (CVDR) to successfully quantify critical directional imbalances that traditional aggregate v/c analysis fails to resolve.

Abstract: Railway transportation of hazardous materials (HazMat) exhibits a characteristic “low-probability, high-consequence” risk profile, compounded by spatially inequitable risk distribution across the network and increasingly stringent carbon emission constraints. Achieving synergistic optimization across safety, equity, and low-carbon dimensions constitutes a critical and unresolved scientific challenge. To address this, we formulate a robust multi-objective path optimization model (MORPO) that simultaneously minimizes three objectives: Conditional Value-at-Risk (CVaR) to capture extreme tail accident risk, the Gini coefficient to quantify regional risk-allocation equity, and traction-energy carbon emissions to represent ecological impact, all under freight-volume uncertainty. The Bertsimas–Sim robust counterpart theory is employed to equivalently transform the nonlinear uncertain constraints into deterministic linear constraints, ensuring computational tractability. To tackle the high-dimensional discrete nature, strong multi-objective conflicts, and non-convex Pareto-front characteristics of the problem, we propose an Adaptive Crossover-Mutation and Elite-preservation NSGA-II algorithm (ACE-NSGA-II), which integrates dual-strategy initialization, individual-level adaptive crossover and mutation operators, and a hierarchical elite preservation mechanism, thereby overcoming the premature convergence and front-degradation limitations of classical algorithms on three-dimensional non-convex fronts. The approach is validated on a representative North China railway freight network comprising 30 hub nodes and 50 mainline sections. Results demonstrate that: (1) ACE-NSGA-II significantly outperforms NSGA-II, NSGA-III, MOEA/D, and SPEA2 across IGD, HV, and Spread metrics, achieving a 58.7% reduction in IGD and a 23.7% increase in HV relative to standard NSGA-II; (2) under five uncertainty disturbance scenarios, the robust model yields a 13.3%–21.4% improvement in CVaR over the deterministic model, with larger advantages under stronger disturbances; (3) ablation experiments confirm that dual-strategy initialization and hierarchical elite preservation are the two most impactful components, and the four innovations exhibit significant positive synergistic effects; (4) parameter sensitivity analysis reveals that moderate robust conservatism (Γ∈[5,7]) paired with an appropriate population size (N∈[100,150]) achieves three-dimensional synergistic optimality. This work provides both a theoretical model and an algorithmic tool for safety–equity–low-carbon coordinated decision-making in railway hazardous materials transportation.

Abstract: Sustainable and demand-adaptive railway operation requires frameworks capable of aligning service capacity with time-varying demand while ensuring safe, operationally feasible, and energy-efficient service. This paper proposes an integrated control-and-planning framework for modular pod-based railway operation based on virtual coupling. The framework combines a convoy control layer, which ensures safe and dynamically feasible virtually coupled operation, with a planning layer formulated as a mixed-integer linear programming (MILP) model for daily service allocation and convoy sizing. This hierarchical automation architecture coordinates real-time convoy control with service-level planning to adapt capacity to passenger demand. The proposed methodology is evaluated through comparative simulations under peak-hour, shoulder-period, and off-peak demand scenarios, as well as over a daily schedule of 20 services. Its performance is compared with a conventional fixed-composition diesel-electric multiple unit (DEMU)-based operation. Results show that the pod-based configuration increases energy consumption under peak-hour conditions, remains comparable during shoulder periods, and substantially reduces energy consumption in off-peak operation, achieving a 57% saving in that regime. At the daily level, total energy consumption decreases from 3864 kWh to 3075 kWh, corresponding to a 20% reduction. These findings indicate that the main value of the proposed framework lies in transforming convoy composition into a demand-adaptive operational variable, thereby improving energy performance at the daily system level while preserving the safe and dynamically feasible operation of virtually coupled pod formations.

Abstract: The rise of urban air mobility (UAM) has made vertiport siting a critical planning problem linking ground and air transport. Conventional location studies typically treat demand as exogenous, capacity as a hard constraint, and vertiports as always available, thereby failing to capture demand feedback, in-facility congestion, and weather disruptions in real operations. This paper develops a demand-endogenous, capacity-coupled, distributionally robust six-objective mixed-integer nonlinear programming model. In-facility delay under graded capacity is characterized by M/M/c queues; a multinomial logit mode-choice mechanism lets demand evolve endogenously with the layout, forming a demand-congestion fixed-point coupling; and a Wasserstein distributionally robust framework adds a worst-case serviceable-reliability objective. To solve the six-dimensional many-objective problem, a reference-point-based improved NSGA-III is designed, featuring a three-layer nested evaluation (outer evolution-middle fixed point-inner robust dual) with evaluation caching and vectorized acceleration (a surrogate model is retained as an optional extension). A case study of Huangpu District, Guangzhou, shows that the proposed algorithm attains the best hypervolume (the combined convergence-diversity metric) at the lowest computational cost, with IGD+ comparable to NSGA-II and superior to standard NSGA-III and MOEA/D. The model further internalizes the self-limiting effect of in-facility congestion on hub over-concentration (a congestion-coverage tension) and quantifies the price of robustness: the guaranteeable worst-case serviceable reliability decreases monotonically as the uncertainty budget grows while nominal coverage remains saturated; it also reveals the significant influence of endogenous demand feedback on the UAM-captured demand and in-facility congestion.

Abstract: Range anxiety remains a barrier to the adoption of electric vehicles (EVs), compounded by the energy draw of Heating, Ventilation, and Air Conditioning (HVAC) systems, which can reduce battery reserves by over 20%. Existing data-driven energy estimation models frequently operate as uninterpretable "black boxes", exhibit data leakage during temporal cross-validation, and lack real-time feedback mechanisms for the driver. To address these limitations, this paper presents a physics-informed Human-Machine Interface (HMI) framework driven by a decoupled hybrid machine learning architecture. The proposed layout separates powertrain dynamics from cabin thermal behavior using a multi-engine design: a HistGradientBoosting regressor integrated with a mechanistic Vehicle Specific Power (VSP) indicator to track traction demand, and a regularized Random Forest model enriched with a Newtonian thermal decay function to predict transient thermodynamic climate states. The models were trained and evaluated using a 5-Fold GroupKFold validation protocol across an empirical 55-trip dataset to prevent data leakage and ensure out-of-sample generalization. The traction subsystem achieved an out-of-sample coefficient of determination (R² = 0.9869, Mean Absolute Error MAE = 0.71 kW), while the thermal core reached an R² = 0.8545 (MAE = 0.30 kW). Model transparency and physical consistency were evaluated using game-theoretic SHAP (SHapley Additive exPlanations) values. Finally, a multi-objective Pareto optimization loop was integrated within a real-time HMI advisory dashboard to balance range extension against passenger thermal discomfort. Macroscopic simulation across the empirical fleet dataset indicates that implementing this software-only, context-aware advisory system reduces total fleet grid-charging demand by 5.21%.