Automatic Dependent Surveillance-Broadcast (ADS-B) systems provide continuous aircraft position reports for aviation surveillance and short-term trajectory prediction. However, many data-driven predictors directly model the longitude, latitude, and altitude contained in ADS-B messages as generic multivariate time-series variables, which can weaken latitude-dependent displacement, bearing, and local motion relationships during multi-step forecasting. This paper proposes Geometry-Informed Adaptive Time-Series Fusion (GeoATF), a forecasting framework for short-term aircraft trajectory prediction from ADS-B position reports. For multi-route learning, each input window is mapped from global absolute coordinates to a local geodesic chart anchored at the last observation. The predicted local offsets are subsequently decoded into World Geodetic System 1984 (WGS-84) coordinates through great-circle forward navigation. GeoATF combines a patch-based time-series Transformer (PatchTST) temporal branch for modeling historical ADS-B position sequences with a geometric–kinematic branch that constructs great-circle distance, bearing, velocity component, and trajectory change rate features from the same sequence. An adaptive branch weight fusion module then learns horizon-dependent geometric–kinematic branch weights for residual correction. Experiments on 945 ADS-B flights from ten routes demonstrate that GeoATF delivers more accurate multi-horizon trajectory forecasts than the evaluated baselines, with more pronounced advantages at longer forecast horizons. Flight-level statistical analysis supports the robustness of these improvements, while resource evaluation indicates a moderate computational footprint under the common inference protocol. These results suggest that explicitly representing geometric–kinematic information and adaptively fusing it with temporal features can improve short-term aircraft trajectory prediction using only historical ADS-B position data.

Aircraft trajectory prediction supports trajectory-based operations and air traffic management by estimating future aircraft motion. This systematic review synthesizes studies on trajectory, position, and flight state prediction for civil, manned, fixed-wing aircraft during airborne operation from takeoff to touchdown. Following evidence assessment, 198 studies were retained as primary evidence, of which 55 met the predefined criteria for formal performance synthesis. The review uses a Preparation, Prediction, Update, and Output (PPUO) framework and organizes the evidence around modeling methodology, task formulation, and performance evaluation. The findings show that physics-based, state estimation, data-driven, and hybrid methods have different modeling assumptions and operating conditions, and the available evidence does not support any one paradigm as universally superior across tasks and operating scenarios. Reported performance is jointly influenced by prediction horizon, target cardinality, interaction context, and evaluation conditions. The evidence is concentrated on single-target prediction; quantitative evidence for joint multi-aircraft prediction and probabilistic uncertainty evaluation remains limited, while heterogeneity in experimental conditions restricts direct cross-study comparison. Future research should establish reproducible benchmarks with complete reporting of experimental conditions and conduct cross-scenario validation while evaluating trajectory accuracy, uncertainty quality, computational efficiency, and operational assurance together.

Unmanned Aerial Vehicle (UAV), particularly rotary-wing platforms such as quadcopters and octocopters, has evolved from controlled remote sensing platforms into autonomous agents capable of active task execution. This evolution from collect-then-analyze workflows to closed-loop perception, reasoning, and action signifies a paradigm shift toward Embodied AI, unlocking opportunities for the low-altitude economy. However, current research on UAV Embodied AI (UAV-EAI) often implicitly frames the field as a direct extension of indoor robotics or autonomous driving, which overlooks the fundamental distinctions of aerial agents. To bridge this gap, we introduce a comparative framework contrasting UAV-EAI with Indoor-EAI and Autonomous Driving Embodied AI (AD-EAI). By systematically decomposing the domain into nine key dimensions, we (i) analyze core tasks such as perception, localization, and exploration; (ii) review enabling infrastructure, including simulators and datasets; and (iii) categorize modeling methods ranging from physics-centric control to cognition-centric models. Our analysis demonstrates that the convergence of 6-DoF motion space, kilometer-scale unstructured environments, and stringent on-device constraints establishes a research regime qualitatively different from ground-based agents. These factors significantly impede the migration of existing VLM/LLM-based embodied systems for UAVs. Finally, we summarize open challenges and outline promising directions for the next generation of UAV-EAI.

Accurate reconstruction of digital surface models (DSMs) from multi-date high-resolution urban satellite imagery supports large-scale Earth observation and three-dimensional city modeling, with 3D Gaussian Splatting (3DGS) emerging as an efficient solution. Existing shadow-aware methods provide shadow-related spatial evidence, whereas geometric priors are generally applied across the image or scene without using this evidence to allocate supervision. Consequently, regions where shadows overlap abrupt elevation transitions may remain geometrically under-constrained, producing spatially non-uniform elevation errors whose severity varies with the geometry covered by shadow. To address this issue, we propose Shadow-Adaptive Geometric Supervision for Earth Observation Gaussian Splatting (SAGE). SAGE is primarily designed for urban scenes dominated by regular human-made structures and retains image-wide altitude guidance while allocating additional geometric supervision at two levels. At the spatial level, patch-level shadow weights determine where a shadow-targeted local altitude prior is applied to constrain relative altitude agreement. At the scene level, SAGE computes the reconstruction-derived Shadow Geometric Complexity (SGC) from multi-view shadow evidence and gradient statistics of the predicted DSM. SGC is used as a scene-level descriptor of shadow–gradient contrast and determines how strongly the local altitude prior and the complementary Manhattan orientation regularization are applied. Experiments on seven predominantly urban scenes from the DFC2019 and IARPA 2016 benchmarks show that SAGE achieves the best mean DSM accuracy on both benchmarks while retaining minute-level processing.

Geospatial target localization refers to determining the geographic location of targets from natural language queries, which is essential for applications such as urban management, emergency response, and location-based services. As user queries evolve from address matching toward complex forms incorporating multiple spatial semantic factors, the localization process increasingly demands the coordination of heterogeneous tools and data sources. Automation remains hindered by the lack of a unified structured representation for semantic factors, the difficulty of translating complex spatial constraints into executable operations, and the unreliability of single-round planning and execution. This study proposes SGTAgent, an agent framework for geospatial target localization from complex spatial semantic queries. The parser transforms queries into structured spatial semantic factors, the planner organizes dependent spatial workflows, and the feedback mechanism distinguishes local parameter refinement from workflow restructuring. Experiments show that SGTAgent achieves higher localization accuracy than the evaluated generic baselines across multiple LLM backbones. Using DeepSeek-V4.1-Flash, SGTAgent achieves localization accuracies of 39.86%, 68.76%, and 81.88% at distance thresholds of 100, 300, and 500 m, respectively, outperforming the evaluated agents under the corresponding experimental configurations. On complex queries using GPT-4o, it achieves an LA@500m of 80.47%, compared with 70.12% for ReAct. Ablation experiments show that removing the complete Restate–Replan feedback mechanism reduces the 500 m localization accuracy from 81.88% to 74.83%, demonstrating the effectiveness of the execution feedback mechanism under the evaluated configuration. An independent evaluation of the parser achieves macro-F1 scores of 89.76% on the original benchmark and 89.62% on independently formulated queries, indicating consistent semantic factor extraction capability across different query formulations.