Autonomous navigation is a core enabler of smart agriculture, where path planning and trajectory tracking control play essential roles in achieving efficient and precise operations. Path planning determines operational efficiency and coverage completeness, while trajectory tracking directly affects task accuracy and system robustness. This paper presents a systematic review of agricultural robot navigation research published between 2020 and 2025, based on literature retrieved from major databases including Web of Science and EI Compendex (ultimately including 95 papers). Research advances in global planning (coverage and point-to-point), local planning (obstacle avoidance and replanning), multi-robot cooperative planning, and classical, advanced, and learning-based trajectory tracking control methods are comprehensively summarized. Particular attention is given to their application and limitations in typical agricultural scenarios such as open-fields, orchards, greenhouses, and hilly slopes. Despite notable progress, key challenges remain, including limited algorithm comparability, weak cross-scenario generalization, and insufficient long-term validation. To address these issues, a scenario-driven “scenario–constraint–performance” adaptive framework is proposed to systematically align navigation methods with environmental and operational conditions, providing practical guidance for developing scalable and engineering-ready agricultural robot navigation systems.

ABSTRACT There is increasing impetus to help individuals navigate health and social care systems. Though the development and effectiveness of navigator roles have long been explored in health and sociological research, evaluations of navigation are typically narrow in focus and oriented to individual outcomes, quantitative metrics, or barriers and enablers in implementation. Investigating the structural effects of navigation, especially across disparate contexts and sectors, presents significant challenges for researching navigation. To explore these challenges, this article presents a critical interpretive synthesis of qualitative literature on navigation spanning health and social care. Twenty qualitative studies were included, with analysis organised across four themes: (1) modalities of navigation practice; (2) epistemic authority and professional identity; (3) authorising navigation in and through place; (4) situating navigation and its effects in systems. We conceptualise navigation as operating via dual modalities of structural and interpretative practice, and argue this conceptualisation facilitates closer critical attention to the relational and situated practices often obscured in accounts of navigators’ work. We also highlight a need for strengthened research designs that reflect the complexities of care systems and consider the effects of navigation across multiple sectors. We finally reflect on emerging challenges posed by digital and algorithmic tools in navigation.

Ensuring navigational safety is one of the most critical challenges in autonomous maritime navigation research, requiring accurate real-time assessment of collision risks and prompt navigational decisions based on such assessments. Traditional rule-based systems utilizing radar and Automatic Identification Systems (AIS) exhibit fundamental limitations in simultaneously analyzing discrete objects such as vessels and buoys alongside continuous environmental boundaries like coastlines and bridges. To address these limitations, recent research has incorporated artificial intelligence approaches, though most recent studies have primarily focused on object detection methods. This study proposes a structured tag-based multimodal navigation safety framework that performs inference on maritime scenes by integrating YOLO-based object detection with the LLaVA vision–language model, generating outputs that include risk level assessment, navigation action recommendations, reasoning explanations, and object information. The proposed method achieved 86.1% accuracy in risk level assessment and 76.3% accuracy in navigation action recommendations. Through a hierarchical early stopping system using delimiter-based tags, the system reduced output token generation by 95.36% for essential inference results and 43.98% for detailed inference results compared to natural language outputs.

When we drive into a supermarket, navigate in a large shopping mall or find our way in an unfamiliar city, we utilize our attention and spatial memory to reach our goal. Previous research shows that the integration of memory, perception, and executive functions is essential for efficient navigation, and its capacity can vary greatly among individuals. It is unclear what specific factors drive individual differences in navigation abilities and spatial cognition. This study investigates the neural and behavioral mechanisms underlying landmark and map-based learning during goal-directed navigation. Our main contributions in this project contain two parts: 1) up to our best knowledge, I successfully developed the first novel goal-directed navigation research platform, NeuroNav, for investigating the neural mechanism in navigation. The maze design of NeuroNav is inspired by Tolman's multiple T-junctions. During testing, participants can observe the environment in the maze via a built-in camera and control their view and location via screw-driven linear slides, a gamepad controller, and an Arduino microcontroller; 2) I systematically examined how individual differences in attention, working memory, and spatial familiarity influence navigation performance. Participants navigated to different goal locations both with and without a map while behavioral metrics (e.g., time to goal, heading changes) and EEG signals (concentration, theta, beta, gamma bands) were recorded. Results showed that familiarity and map use significantly improved navigation efficiency and reduced cognitive load, as reflected in both behavior and neural activity. EEG recordings revealed increased theta and gamma activity during novel landmark encoding and decision-making phases. These findings highlight the interplay between attention, memory, and environmental cues in spatial learning, with implications for assistive navigation technologies and populations with spatial memory deficits.

This study presents a comparative analysis of the Dijkstra and A* algorithms for the autonomous path planning of Unmanned Aerial Vehicles (UAVs) in simulated 2D environments. The simulations were conducted in CoppeliaSim (V-REP), a versatile robotics simulation platform, where a quadcopter model navigated through obstacle-rich scenarios by following the shortest path generated by each algorithm. Both algorithms were implemented using a grid-based graph representation, with the path costs calculated using the Manhattan and Euclidean distances. The UAV visually traced the computed path in real time, avoided obstacles, and returned to the starting point after reaching the target. Performance metrics such as path optimality, computational efficiency, and execution time were evaluated to compare the two approaches. The results indicate that while Dijkstra guarantees the shortest path, A* achieves faster convergence with minimal deviation in path length, making it more suitable for real-time UAV navigation. The visualized simulation framework demonstrates the effectiveness of integrating classical pathfinding algorithms with UAV models in a physics-enabled environment, offering a reproducible testbed for autonomous navigation research.

Graph properties drive navigational selection between equidistant routes.

Place recognition—the ability to identify previously visited locations—is critical for both biological navigation and autonomous systems. This review synthesizes findings from robotic systems, animal studies, and human research to explore how different systems encode and recall place. We examine the computational and representational strategies employed across artificial systems, animals, and humans, highlighting convergent solutions such as topological mapping, cue integration, and memory management. Animal systems reveal evolved mechanisms for multimodal navigation and environmental adaptation, while human studies provide unique insights into semantic place concepts, cultural influences, and introspective capabilities. Artificial systems showcase scalable architectures and data-driven models. We propose a unifying set of concepts by which to consider and develop place recognition mechanisms and identify key challenges such as generalization, robustness, and environmental variability. This review aims to foster innovations in artificial localization by connecting future developments in artificial place recognition systems to insights from both animal navigation research and human spatial cognition studies.

This article presents the rationale and design for the adaptation and implementation of a patient navigation program for cervical cancer screening across contexts in Senegal. A model, based on the NIH NCI Patient Navigator Research Program (PNRP) model, informs the proposed program for adaptation which aims to reduce intrapersonal- (knowledge, communication), interpersonal- (stigma, misinformation), and community-level (women’s lack of autonomy in healthcare decision-making) barriers. The specific aims of the study are to: 1) Evaluate the adaptation process of the evidence-based Patient Navigation Model utilizing the Dynamic Adaptation Process (DAP) across rural and urban contexts in Kedougou and Dakar, Senegal; 2) Conduct an effectiveness-implementation hybrid type 1 stepped-wedge randomized pragmatic trial of the adapted patient navigation program across Kedougou and Dakar, Senegal, and 3) Evaluate the implementation outcomes (feasibility, acceptability, fidelity, penetrance, sustainability, and cost) of The Adapted Patient Navigation Program across multiple contexts in the Kedougou and Dakar regions, using mixed methods and guided by the Exploration, Preparation, Implementation, Sustainment (EPIS) Framework. The Adapted Program is integrated into the existing community health system and is being administered by the Heads of Reproductive Health at the Regional-Level and District Levels who act as Patient Navigator Leaders with oversight by the Regional and District Directors of Health. These individuals coordinate the patient navigation field activities that occur at the health post level. The Community Health Workers (Patient Navigators) are essential to engaging individual clients through education, empowerment, and by accompanying them to the clinical setting for screening and follow-up. The study is a mixed-methods study that collects data from three participant samples: (1) system and organizational stakeholders, (2) patient navigator team members, and (3) clients. The study informs the adaptation and implementation of patient navigation programs for cervical cancer screening in Senegal and other low- and middle-income countries.

Direct optical path measurement in fibre-optic gyroscopes: A potential method for compensating slow-drifting errors

The cognitive mechanisms of spatial perspective taking in map reading.

Pilot feasibility of a financial and health-related social needs navigation intervention (AYA-NAV) for adolescents and young adults with Cancer: Study protocol for a prospective, single-arm study.

This paper describes an innovative wireless mobile robotics control system based on speech recognition, where the ESP32 microcontroller is used to control motors, facilitate Bluetooth communication, and deploy an Android application for the real-time speech recognition logic. With speech processed on the Android device and motor commands handled on the ESP32, the study achieves significant performance gains through distributed architectures while maintaining low latency for feedback control. In experimental tests over a range of 1–10 m, stable 110–140 ms command latencies, with low variation (±15 ms) were observed. The system’s voice and manual button modes both yield over 92% accuracy with the aid of natural language processing, resulting in training requirements being low, and displaying strong performance in high-noise environments. The novelty of this work is evident through an adaptive keyword spotting algorithm for improved recognition performance in high-noise environments and a gradual latency management system that optimizes processing parameters in the presence of noise. By providing a user-friendly, real-time speech interface, this work serves to enhance human–robot interaction when considering future assistive devices, educational platforms, and advanced automated navigation research.

Spinal navigation systems improve pedicle screw placement accuracy, but their reliance on supine preoperative imaging can introduce errors due to positional differences between preoperative and intraoperative spinal alignment. These misalignments may compromise surgical outcomes, particularly in lumbar spine procedures. This study investigates how key lumbar and lumbopelvic parameters differ between prone and supine positions, aiming to refine imaging workflows and surgical navigation practices. A retrospective cohort study analyzed paired prone and supine CT images from 85 adult patients in the ACRIN database. Key parameters-pelvic tilt, lumbar lordosis, L1 slope, pelvic incidence, and L1-L5 Cobb angle-were measured. Statistical significance was assessed using two-tailed t-tests, with pairwise comparisons conducted to evaluate positional differences. Significant differences were observed in pelvic tilt (mean prone-supine difference: 4.27°, p = 0.0002) and L1 slope (mean prone-supine difference: 3.16°, p = 0.001). Other parameters, including lumbar lordosis, pelvic incidence, and L1-L5 Cobb angle, showed no significant differences. Our study provides the first comprehensive analysis of prone versus supine alignment in the lumbar spine, addressing a critical gap in spinal navigation research. The findings underscore the limitations of supine preoperative imaging in reflecting intraoperative conditions. Incorporating these insights into navigation workflows can improve registration accuracy and surgical outcomes. Future innovations, such as AI-based predictive modeling, may further address positional discrepancies and optimize lumbar spine surgeries. This work highlights the importance of advancing imaging protocols to align with intraoperative realities.

How learning generalizes across contexts is a fundamental question inneuroscience, with broad implications for adaptive behavior and cognition. The hippocampus(HPC) plays a key role in contextual learning, but HPC cells exhibit place-specificactivity that reorganizes, or ‘remaps’ across environments, raising thequestion of how stable, task-relevant representations can be preserved. Here, we usedcalcium imaging to monitor hippocampal neuron activity as rats performed a conditioningtask across multiple spatial contexts. We asked whether hippocampal neurons, which encodeboth spatial locations and task-relevant features, could maintain stable representationsof the task despite remapping of spatial codes. To assess representational consistency, weapplied dimensionality reduction and machine learning to construct manifold embeddings ofpopulation-level HPC activity. We found that task-related neural representations remainedstable across different environments, even as spatial representations shifted. Moreover,these representations exhibited similar geometric structure not only across contextswithin individual animals, but also across different animals, suggesting the presence of ashared neural syntax for associative learning in the hippocampus. These findings bridge acritical gap between memory and navigation research, revealing how stable cognitiverepresentations emerge from dynamic spatial codes. They provide new insight into conservedhippocampal encoding strategies, with potential relevance for understanding flexiblememory, learning, and their disruption in neuropsychiatric disorders.

For the vast majority of spatial navigation research, experimental tasks are implemented in real-world environments. In recent decades, there has been an increasing trend toward virtual environments, which offer several benefits compared to their real-world counterparts while also having certain limitations. With these properties in mind, we have developed the Gallery of Memories (GA-ME), a customizable virtual-navigation task that is equipped for the assessment of both spatial navigation and memory within a highly controlled three-dimensional environment. The GA-ME provides a 3D position and head direction (pitch and yaw) sampling rate that is significantly higher compared to alternatives, enabling users to reconstruct a participant’s movement in the environment with remarkable spatiotemporal precision while its design, including nested spaces, makes it optimal for the study of place and grid cells in humans. These properties imbue the GA-ME with the potential to be widely utilized in both research and clinical settings for the in-depth study of spatial navigation and memory, with the possibility of conducting human intra- and extra-cranial electrophysiology, imaging, and eye-tracking measurements relevant to these faculties.

Visuospatial function is a critical aspect of cognitive abilities, encompassing visual perception, attention, memory, and adaptive responses to spatial changes. This paper reviews studies on human visuospatial function, spatial navigation, and factors contributing to visuospatial impairments. After introducing fundamental concepts of visuospatial function and spatial navigation, classical methods for assessing visuospatial performance are summarized. By examining recent advances in spatial navigation studies, this paper discusses factors influencing spatial navigation capabilities and explores how spatial navigation paradigms can be used to investigate visuospatial cognitive impairments. Finally, current limitations in spatial navigation research are highlighted. Overall, the current research has not yet reached definitive conclusions regarding visuospatial aspects. However, this paper aims to enhance the understanding of visuospatial dysfunction and spatial navigation, providing valuable references for future research.

This study aimed to enhance the object recognition capabilities of autonomous vehicles in constrained and dynamic environments. By integrating Light Detection and Ranging (LiDAR) technology with a modified Voxel-RCNN framework, the system detected and classified six object classes: human, wall, car, cyclist, tree, and cart. This integration improved the safety and reliability of autonomous navigation. The methodology included the preparation of a point cloud dataset, conversion into the KITTI format for compatibility with the Voxel-RCNN pipeline, and comprehensive model training. The framework was evaluated using metrics such as precision, recall, F1-score, and mean average precision (mAP). Modifications to the Voxel-RCNN framework were introduced to improve classification accuracy, addressing challenges encountered in complex navigation scenarios. Experimental results demonstrated the robustness of the proposed modifications. Modification 2 consistently outperformed the baseline, with 3D detection scores for the car class in hard scenarios increasing from 4.39 to 10.31. Modification 3 achieved the lowest training loss of 1.68 after 600 epochs, indicating significant improvements in model optimization. However, variability in the real-world performance of Modification 3 highlighted the need for balancing optimized training with practical applicability. Overall, the study found that the training loss decreased up to 29.1% and achieved substantial improvements in detection accuracy under challenging conditions. These findings underscored the potential of the proposed system to advance the safety and intelligence of autonomous vehicles, providing a solid foundation for future research in autonomous navigation and object recognition.

High-precision localization is critical for intelligent robotics in autonomous driving, smart agriculture, and military operations. While Global Navigation Satellite System (GNSS) provides global positioning, its reliability deteriorates severely in signal degraded environments like urban canyons. Cross-view pose estimation using aerial-ground sensor fusion offers an economical alternative, yet current datasets lack field scenarios and high-resolution LiDAR support.This work introduces a multimodal cross-view dataset addressing these gaps. It contains 29,940 synchronized frames across 11 operational environments (6 field environments, 5 urban roads), featuring: 1) 144-channel LiDAR point clouds, 2) ground-view RGB images, and 3) aerial orthophotos. Centimeter-accurate georeferencing is ensured through GNSS fusion and post-processed kinematic positioning. The dataset uniquely integrates field environments and high-resolution LiDAR-aerial-ground data triplets, enabling rigorous evaluation of 3-DoF pose estimation algorithms for orientation alignment and coordinate transformation between perspectives.This resource supports development of robust localization systems for field robots in GNSS-denied conditions, emphasizing cross-view feature matching and multisensor fusion. Light Detection And Ranging (LiDAR)-enhanced ground truth further distinguishes its utility for complex outdoor navigation research.

Indoor Pedestrian Navigation Research Based on Zero Velocity Correction and Sliding Window

Navigating uneven, unstructured terrain with dynamic obstacles remains a challenge for autonomous mobile robots. This article introduces Dynamic Unstructured Environment (DUnE) for evaluating the performance of off-road navigation systems in simulation. DUnE is a versatile software framework that implements the Gymnasium reinforcement learning (RL) interface for ROS 2, incorporating unstructured Gazebo simulation environments and dynamic obstacle integration to advance off-road navigation research. The testbed automates key performance metric logging and provides semi-automated trajectory generation for dynamic obstacles including simulated human actors. It supports multiple robot platforms and five distinct unstructured environments, ranging from forests to rocky terrains. A baseline reinforcement learning agent demonstrates the framework’s effectiveness by performing pointgoal navigation with obstacle avoidance across various terrains. By providing an RL interface, dynamic obstacle integration, specialized navigation tasks, and comprehensive metric tracking, DUnE addresses significant gaps in existing simulation tools.