Navigation Decisions Research Articles

An autonomous navigation system with a trajectory prediction-based decision mechanism for rubber forest navigation

The use of rubber-tapping robots capable of autonomous navigation in place of manual rubber-tapping is a growing trend, but the challenging multi-objective navigation task in forest environments impedes their autonomous operation. To tackle this issue, an autonomous navigation system with a trajectory prediction-based decision mechanism for rubber forest navigation is designed. This navigation decision mechanism is comprised of obtaining coordinates of target points (OCTP), selecting the next coordinate (SNC), generating the additional coordinates (GAC), and optimizing the planned paths (OPP). By utilizing this mechanism, the robot can autonomously select the next target point based on its current position and the actual operating logic while navigating in the forest areas, adding additional coordinates during row or column changes, and planning and optimizing the path. The on-site experiments demonstrate that during autonomous navigation, the positioning accuracy is favorable and supports subsequent operations. The overall rationality of the planned path reaches 92.14%, further confirming its effectiveness.

Design and Testing of a Tractor Automatic Navigation System Based on Dynamic Path Search and a Fuzzy Stanley Model

Design and Testing of a Tractor Automatic Navigation System Based on Dynamic Path Search and a Fuzzy Stanley Model

Learning Heterogeneous Relation Graph and Value Regularization Policy for Visual Navigation.

The goal of visual navigation is steering an agent to find a given target object with current observation. It is crucial to learn an informative visual representation and robust navigation policy in this task. Aiming to promote these two parts, we propose three complementary techniques, heterogeneous relation graph (HRG), a value regularized navigation policy (VRP), and gradient-based meta learning (ML). HRG integrates object relationships, including object semantic closeness and spatial directions, e.g., a knife is usually co-occurrence with bowl semantically or located at the left of the fork spatially. It improves visual representation learning. Both VRP and gradient-based ML improve robust navigation policy, regulating this process of the agent to escape from the deadlock states such as being stuck or looping. Specifically, gradient-based ML is a type of supervision method used in policy network training, which eliminates the gap between the seen and unseen environment distributions. In this process, VRP maximizes the transformation of the mutual information between visual observation and navigation policy, thus improving more informed navigation decisions. Our framework shows superior performance over the current state-of-the-art (SOTA) in terms of success rate and success weighted by length (SPL). Our HRG outperforms the Visual Genome knowledge graph on cross-scene generalization with ≈ 56% and ≈ 39% improvement on Hits@ 5* (proportion of correct entities ranked in top 5) and MRR * (mean reciprocal rank), respectively. Our code and HRG datasets will be made publicly available in the scientific community.

Vision-and-language navigation based on history-aware cross-modal feature fusion in indoor environment

Vision-and-language navigation (VLN) is a challenging task that requires an agent to navigate an indoor environment using natural language instructions. Traditional VLN employs cross-modal feature fusion, where visual and textual information are combined to guide the agent’s navigation. However, incomplete use of perceptual information, scarcity of domain-specific training data, and diverse image and language inputs result in suboptimal performance. Herein, we propose a cross-modal feature fusion VLN history-aware information, that leverages an agent’s past experiences to make more informed navigation decisions. The regretful model and self-monitoring models are added, and the advantage actor critic(A2C) reinforcement learning algorithm is employed to improve the navigation success rate, reduce action redundancy, and shorten navigation paths. Subsequently, a data augmentation method based on speaker data is introduced to improve the model generalizability. We evaluate the proposed algorithm on the room-to-room (R2R) and room-for-room (R4R) benchmarks, and the experimental results demonstrate that, by comparison, the proposed algorithm outperforms state-of-the-art methods.

Approach angle estimation method for ships based on deep learning

Accurately estimating the ship's approach angle is crucial during berthing, or passing through navigational structures. It forms the foundation for ship navigation decisions. This study introduces the LiDAR point cloud-based ship pose perception network (PP-Net), a novel deep learning-based method for acquiring ship approach angles. Addressing the challenge of manually annotating approach angles, the ship pose perception method based on point cloud feature distribution (FD) was proposed to obtain initial angle estimates. The max-pooling function was utilized to address the disorder of point clouds, and a multi-resolution feature extractor was employed to enhance the network's robustness in extracting features from inconsistently dense point clouds. By using iterative farthest point sampling (IFPS), datasets with varying numbers of point clouds were standardized to a fixed number of points, thereby normalizing input dimensions and improving computational efficiency. Additionally, a novel loss function was proposed to improve the model's predictive accuracy. Experimental validation conducted in the scenario of ships entering ship lifts demonstrates the effectiveness of the proposed method. The outcomes can support safe navigation for ships in restricted waterways.

Adaptive closed-loop modulation of cortical theta oscillations: Insights into the neural dynamics of navigational decision-making

Navigational decision-making tasks, such as spatial working memory (SWM), rely highly on information integration from several cortical and sub-cortical regions. Performance in SWM tasks is associated with theta rhythm, including low-frequency oscillations related to movement and memory. The interaction of the ventral hippocampus (vHPC) and medial prefrontal cortex (mPFC), reflected in theta synchrony, is essential in various steps of information processing during SWM. We used a closed-loop neurofeedback (CLNF) system to upregulate theta power in the mPFC and investigate its effects on circuit dynamics and behavior in animal models. Specifically, we hypothesized that enhancing the power of the theta rhythm in the mPFC might improve SWM performance. Animals were divided into three groups: closed-loop (CL), random-loop (RL), and OFF (without stimulation). We recorded local field potential (LFP) in the mPFC while electrical reward stimulation contingent on cortical theta activity was delivered to the lateral hypothalamus (LH), which is considered one of the central reward-associated regions. We also recorded LFP in the vHPC to evaluate the related subcortical neural changes. Results revealed a sustained increase in the theta power in both mPFC and vHPC for the CL group. Our analysis also revealed an increase in mPFC-vHPC synchronization in the theta range over the stimulation sessions in the CL group, as measured by coherence and cross-correlation in the theta frequency band. The reinforcement of this circuit improved spatial decision-making performance in the subsequent behavioral results. Our findings provide direct evidence of the relationship between specific theta upregulation and SWM performance and suggest that theta oscillations are integral to cognitive processes. Overall, this study highlights the potential of adaptive CLNF systems in investigating neural dynamics in various brain circuits.

A Dynamic Bayesian Network model for ship navigation risk in the Arctic Northeast Passage

The melting of a significant amount of sea ice has enabled the commercial and normal operation of ships in the Arctic. However, the harsh natural environment, unique geographical location and complex navigational conditions pose a serious challenge to the safety of ships during navigation. Ship besetting in ice and ship-ice collision are identified as the most critical risks, potentially resulting in vessel damage, property loss, environmental harm, and even casualties. This study proposes a dynamic Bayesian network (DBN) model to assess and predict the risk of ship besetting in ice and ship-ice collision in five sea areas along the Arctic Northeast Passage. The proposed method comprehensively considers environmental, ship and human factors and incorporates multi-year (2013–2022) marine meteorological reanalysis data and expert knowledge. The developed model has seven time steps and can be updated based on past, current and future conditions. The results indicate that it can effectively reflect the distribution of ship navigation risks in various sea areas and has the capability to be updated in real-time, serving as a tool for guiding ship operations and navigation decisions. At the same time, this study confirms August and September are the optimal navigation periods for the Northeast Passage, and the East Siberian Sea presenting the highest navigation risk. Ice thickness, ice concentration, navigation experience, professional skills, ship speed and ship class are identified as the primary risk factors in the region.

A conceptual framework on the role of magnetic cues in songbird migration ecology.

Migrating animals perform astonishing seasonal movements by orienting and navigating over thousands of kilometres with great precision. Many migratory species use cues from the sun, stars, landmarks, olfaction and the Earth's magnetic field for this task. Among vertebrates, songbirds are the most studied taxon in magnetic-cue-related research. Despite multiple studies, we still lack a clear understanding of when, where and how magnetic cues affect the decision-making process of birds and hence, their realised migratory behaviour in the wild. This understanding is especially important to interpret the results of laboratory experiments in an ecologically appropriate way. In this review, we summarise the current findings about the role of magnetic cues for migratory decisions in songbirds. First, we review the methodological principles for orientation and navigation research, specifically by comparing experiments on caged birds with experiments on free-flying birds. While cage experiments can show the sensory abilities of birds, studies with free-flying birds can characterise the ecological roles of magnetic cues. Second, we review the migratory stages, from stopover to endurance flight, in which songbirds use magnetic cues for their migratory decisions and incorporate this into a novel conceptual framework. While we lack studies examining whether and when magnetic cues affect orientation or navigation decisions during flight, the role of magnetic cues during stopover is relatively well studied, but mostly in the laboratory. Notably, many such studies have produced contradictory results so that understanding the biological importance of magnetic cues for decisions in free-flying songbirds is not straightforward. One potential explanation is that reproducibility of magnetic-cue experiments is low, probably because variability in the behavioural responses of birds among experiments is high. We are convinced that parts of this variability can be explained by species-specific and context-dependent reactions of birds to the study conditions and by the bird's high flexibility in whether they include magnetic cues in a decision or not. Ultimately, this review should help researchers in the challenging field of magnetoreception to design experiments meticulously and interpret results of such studies carefully by considering the migration ecology of their focal species.

LoCoMoTe - A Framework for Classification of Natural Locomotion in VR by Task, Technique and Modality.

Virtual reality (VR) research has provided overviews of locomotion techniques, how they work, their strengths and overall user experience. Considerable research has investigated new methodologies, particularly machine learning to develop redirection algorithms. To best support the development of redirection algorithms through machine learning, we must understand how best to replicate human navigation and behaviour in VR, which can be supported by the accumulation of results produced through live-user experiments. However, it can be difficult to identify, select and compare relevant research without a pre-existing framework in an ever-growing research field. Therefore, this work aimed to facilitate the ongoing structuring and comparison of the VR-based natural walking literature by providing a standardised framework for researchers to utilise. We applied thematic analysis to study methodology descriptions from 140 VR-based papers that contained live-user experiments. From this analysis, we developed the LoCoMoTe framework with three themes: navigational decisions, technique implementation, and modalities. The LoCoMoTe framework provides a standardised approach to structuring and comparing experimental conditions. The framework should be continually updated to categorise and systematise knowledge and aid in identifying research gaps and discussions.

USV formation navigation decision-making through hybrid deep reinforcement learning using self-attention mechanism

To address the challenging of balancing Unmanned Surface Vessel (USV) autonomous collision avoidance and formation maintenance in uncertain environments, a formation construction and navigation decision-making strategy based on Hybrid Deep Reinforcement Learning (HDRL) is proposed in this study. The novelty of this study is that: (1) A HDRL training approach is proposed, incorporating diverse DRL for virtual leader and followers, thereby significantly enhancing the decision-making adaptability of USVs formation. (2) The multi-head attention mechanism of decentralized Critic strategy is used to enhance the attentional focus of the HDRL algorithm towards different agents, thereby effectively improving convergence speed. (3) The method employs a meticulously designed hybrid reward function and incorporates the Optimal Reciprocal Collision Avoidance (ORCA) for speed selection, thereby providing optimal speed recommendations based on the current situation. It is worth emphasizing that the method exhibits exceptional performance in terms of success rate, navigation time, average reward value, and other relevant indicators within the simulation. Finally, the method is validated through the utilization of real-time navigation data in the Panama Canal, thereby substantiating the potential engineering applicability.

Cognitive issues of mobile map design and use

ABSTRACT Internet-connected mobile devices have changed how people access information. Like other information sources, maps have benefited from and been re-envisioned for mobile devices, and they are used in new contexts. However, these new contexts often generate additional cognitive load. We explored in depth two strategies designers could use to mitigate high cognitive loads associated with mobile map use: offloading cognition and reducing cognitive load by improved design to support the allocation of attention between the map and the environment. In reviewing these strategies, we considered their relevance to several mobile map use cases (navigation, individual and collaborative spatial decision making, information enrichment, and entertainment). Next, we identified recent progress in our understanding of how to measure cognitive load and map use context. Finally, we explored the wider implications of mobile maps for human behaviour and cognition. We identified two important cross-cutting research questions: 1) How can mobile maps be designed to reduce cognitive load by providing what is really needed by users to facilitate their cognitive processes?; 2) How can the intrinsic additional cognitive load created by the characteristics of mobile maps be managed and minimised by supporting the distribution of the user's attention between the map and environment?

Autonomous navigation decision-making for ships in complex estuarine waters: Methodology and validation

This study focused on key scientific issues in the autonomous navigation for ships, including constraints on environmental and navigation rules, limitations on ship maneuverability, and collision avoidance mechanisms, to address the challenges of autonomous navigation decision-making for ships in complex estuarine waters and validate the feasibility and accuracy of the proposed methodology. A novel autonomous navigation decision-making method that can dynamically adapt to residual errors in a system and the random movements of target ships was proposed. The predictive model used to calculate decision plans and the simulation model used to execute and validate decision plans were separated. Environmental constituent elements were classified and modeled to digitize the environment. A navigation scenario was constructed based on the fusion of virtual and real information. The control and process prediction methods were proposed based on the nonlinear ship maneuvering characteristics. Navigation rules were summarized, analyzed, and integrated into the decision-making and execution process. The collision avoidance mechanism and calculation method for feasible course and speed range were studied. Simulation experiments were conducted using the North Channel of the Yangtze River Estuary as an example. The results demonstrated that the proposed method can accurately make autonomous navigation decisions, safely avoid collisions, and timely track routes in nearly realistic scenarios.

Do Magnetic murmurs guide birds? A directional statistical investigation for influence of Earth's Magnetic field on bird navigation.

This paper delves into the intricate relationship between changes in Magnetic inclination and declination at specific geographical locations and the navigational decisions of migratory birds. Leveraging a dataset sourced from a prominent bird path tracking web resource, encompassing six distinct bird species' migratory trajectories, latitudes, longitudes, and observation timestamps, we meticulously analyzed the interplay between these avian movements and corresponding alterations in Magnetic inclination and declination. Employing a circular von Mises distribution assumption for the latitude and longitude distributions within each subdivision, we introduced a pioneering circular-circular regression model, accounting for von Mises error, to scrutinize our hypothesis. Our findings, predominantly supported by hypothesis tests conducted through circular-circular regression analysis, underscore the profound influence of Magnetic inclination and declination shifts on the dynamic adjustments observed in bird migration paths. Moreover, our meticulous examination revealed a consistent adherence to von Mises distribution across all bird directions. Notably, we unearthed compelling correlations between specific bird species, such as the Black Crowned Night Heron and Brown Pelican, exhibiting a noteworthy negative correlation with Magnetic inclination and a contrasting positive correlation with Magnetic declination. Similarly, the Pacific loon demonstrated a distinct negative correlation with Magnetic inclination and a positive association with Magnetic declination. Conversely, other avian counterparts showcased positive correlations with both Magnetic declination and inclination, further elucidating the nuanced dynamics between avian navigation and the Earth's magnetic field parameters.

A Local-and-Global Attention Reinforcement Learning Algorithm for Multiagent Cooperative Navigation.

The cooperative navigation algorithm is the crucial technology for multirobot systems to accomplish autonomous collaborative operations, and it is still a challenge for researchers. In this work, we propose a new multiagent reinforcement learning algorithm called multiagent local-and-global attention actor-critic (MLGA2C) for multiagent cooperative navigation. Inspired by the attention mechanism, we design the local-and-global attention module to dynamically extract and encode critical environmental features. Meanwhile, based on the centralized training and decentralized execution (CTDE) paradigm, we extend a new actor-critic method to handle feature encoding and make navigation decisions. We also evaluate the proposed algorithm in two cooperative navigation scenarios: static target navigation and dynamic pedestrian target tracking. The multiple experimental results show that our algorithm performs well in cooperative navigation tasks with increasing agents.

Social-ecological predictors of spotted hyena navigation through a shared landscape.

Human-wildlife interactions are increasing in severity due to climate change and proliferating urbanization. Regions where human infrastructure and activity are rapidly densifying or newly appearing constitute novel environments in which wildlife must learn to coexist with people, thereby serving as ideal case studies with which to infer future human-wildlife interactions in shared landscapes. As a widely reviled and behaviorally plastic apex predator, the spotted hyena (Crocuta crocuta) is a model species for understanding how large carnivores navigate these human-caused 'landscapes of fear' in a changing world. Using high-resolution GPS collar data, we applied resource selection functions and step selection functions to assess spotted hyena landscape navigation and fine-scale movement decisions in relation to social-ecological features in a rapidly developing region comprising two protected areas: Lake Nakuru National Park and Soysambu Conservancy, Kenya. We then used camera trap imagery and Barrier Behavior Analysis (BaBA) to further examine hyena interactions with barriers. Our results show that environmental factors, linear infrastructure, human-carnivore conflict hotspots, and human tolerance were all important predictors for landscape-scale resource selection by hyenas, while human experience elements were less important for fine-scale hyena movement decisions. Hyena selection for these characteristics also changed seasonally and across land management types. Camera traps documented an exceptionally high number of individual spotted hyenas (234) approaching the national park fence at 16 sites during the study period, and BaBA results suggested that hyenas perceive protected area boundaries' semi-permeable electric fences as risky but may cross them out of necessity. Our findings highlight that the ability of carnivores to flexibly respond within human-caused landscapes of fear may be expressed differently depending on context, scale, and climatic factors. These results also point to the need to incorporate societal factors into multiscale analyses of wildlife movement to effectively plan for human-wildlife coexistence.

RL-TEE: Autonomous Probe Guidance for Transesophageal Echocardiography Based on Attention-Augmented Deep Reinforcement Learning

Ultrasound image acquisition in conventional transesophageal echocardiography (TEE) requires complex manual operation of the probe in the esophagus based on the interpretation of ultrasound images and in-depth knowledge of the cardiac anatomy. In this work, we formulate the TEE probe guidance task as a reinforcement learning (RL) problem, and present the first learning-based solution to 3-DOF control of a TEE probe based on the ultrasound image feedback, named RL-TEE, in order to mimic the visual search and navigation strategies of expert echocardiographers. The probe-tissue interaction in TEE is carefully modeled in our framework by considering both the requirements for navigation towards the standard views and compliance in the esophageal environment. Furthermore, we propose a hybrid deep Q-network model that augments a convolutional neural network backbone with self-attention mechanisms to better capture spatial information in ultrasound images to guide navigation decisions. The presented methods are preliminarily validated in a TEE simulation environment built with data from 25 subjects to acquire four standard views of the heart. Our results show that the proposed method can effectively learn to accurately and compliantly guide the probe movement for TEE standard view acquisition tasks and has a good generalization ability to unseen patient data. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —The motivation of this paper is to realize 3-DOF movement guidance of a TEE probe to acquire the standard views of the heart based on the real-time images, which can be applied to existing robotic control systems or used to assist novice echocardiographers in TEE examination, thereby relieving operator workload and improving ease of use. This paper suggests a novel approach that uses the deep RL technique to achieve automatic interpretation of TEE images and intelligent guidance of the probe movement. The RL framework is designed to take into account both the navigation efficiency and compliance with the esophageal environment for the targeted intracorporeal application. A hybrid deep Q-network model that augments a convolutional neural network with attention mechanisms is designed to better capture spatial information from ultrasound images to predict the probe movement. The effectiveness of the framework is preliminarily validated in extensive experiments in a simulation environment built with real patient data. The proposed method can be applied in clinical use to provide real-time TEE probe guidance for novice echocardiographers, and can be integrated with a robotic system to fully automate the TEE acquisition, thereby relieving the doctors from tedious manual operation to focus on the diagnosis and treatment.

Deep reinforcement learning navigation via decision transformer in autonomous driving.

In real-world scenarios, making navigation decisions for autonomous driving involves a sequential set of steps. These judgments are made based on partial observations of the environment, while the underlying model of the environment remains unknown. A prevalent method for resolving such issues is reinforcement learning, in which the agent acquires knowledge through a succession of rewards in addition to fragmentary and noisy observations. This study introduces an algorithm named deep reinforcement learning navigation via decision transformer (DRLNDT) to address the challenge of enhancing the decision-making capabilities of autonomous vehicles operating in partially observable urban environments. The DRLNDT framework is built around the Soft Actor-Critic (SAC) algorithm. DRLNDT utilizes Transformer neural networks to effectively model the temporal dependencies in observations and actions. This approach aids in mitigating judgment errors that may arise due to sensor noise or occlusion within a given state. The process of extracting latent vectors from high-quality images involves the utilization of a variational autoencoder (VAE). This technique effectively reduces the dimensionality of the state space, resulting in enhanced training efficiency. The multimodal state space consists of vector states, including velocity and position, which the vehicle's intrinsic sensors can readily obtain. Additionally, latent vectors derived from high-quality images are incorporated to facilitate the Agent's assessment of the present trajectory. Experiments demonstrate that DRLNDT may achieve a superior optimal policy without prior knowledge of the environment, detailed maps, or routing assistance, surpassing the baseline technique and other policy methods that lack historical data.

Using a biomimetic sonar bat robot to explore the sensory world of bats in the field and in the laboratory

Bat species such as those belonging to the families of horseshoe and Old World leaf-nosed bats make their navigation decisions based on echoes that are often received from dense vegetation. To gain an understanding of how these animals accomplish navigation based on such complicated clutter echoes, a biomimetic sonar system that replicates the auditory periphery of bats has been used as synthetic observer to probe the sensory stimulus ensemble of the bats in a flight tunnel as well as in the field. Both laboratory and field experiments were carried out in Brunei on the island of Borneo. For the flight tunnel, the biomimetic sonar robot has been used to record a large number of pulse and echo pairs in conjunction with an array consisting of 28 ultrasonic microphones. In this experiment, the sonar head was positioned at points spaced on a 3d grid that covers the tunnel's volume. This data set will be used to train a deep neural network to predict the inputs to the ears of a bat maneuvering in the tunnel. In the field experiments, the sonar head was carried off road in various tropical-forest habitats to record echo samples in conjunction with precise GPS location references.

A framework for improving UAV decision of autonomous navigation from training to application migration under perceptual uncertainty

Unmanned aerial vehicles (UAVs) autonomous navigation based on reinforcement learning usually requires training agents in simulation scenarios and then transferring the trained agents to application scenarios. However, due to serious distribution mismatch between the idealized simulation scenario and the application environment and the inevitable uncertainty perception problem of airborne sensors in complex scenarios, the navigation performance of UAV under migration applications is not ideal. This work fully analyzes the factors that affect UAV navigation performance, including algorithm performance, training strategy, and state awareness. Based on the analysis results, this article proposes a framework to improve the autonomous navigation performance of UAVs in the migration process from training to application, which consists of three parts: ‘scenario-perception-algorithm’. In addition, this paper proposes improvement strategies for each part from the perspectives of spatial features, temporal features, and perceptual denoising. We combine the proposed framework with navigation algorithms to improve the navigation decision-making performance of UAVs in migration applications under uncertainty perception. Many simulation experiments demonstrate the effectiveness of the proposed framework and its robustness to uncertainty perception.

Contribution of the medial entorhinal cortex to performance on the Traveling Salesperson Problem in rats

In order to successfully navigate through space, animals must rely on multiple cognitive processes, including orientation in space, memory of object locations, and navigational decisions based on that information. Although highly-controlled behavioral tasks are valuable for isolating and targeting specific processes, they risk producing a narrow understanding of complex behavior in natural contexts. The Traveling Salesperson Problem (TSP) is an optimization problem that can be used to study naturalistic foraging behaviors, in which subjects select routes between multiple baited targets. Foraging is a spontaneous, yet complex, behavior, involving decision-making, attention, course planning, and memory. Previous research found that hippocampal lesions in rats impaired TSP task performance, particularly on measures of spatial memory. Although traditional laboratory tests have shown the medial entorhinal cortex (MEC) to play an important role in spatial memory, if and how the MEC is involved in finding efficient solutions to the TSP remains unknown. In the current study, rats were trained on the TSP, learning to retrieve bait from targets in a variety of spatial configurations. After recovering from either an MEC lesion or control sham surgery, the rats were tested on eight new configurations. Our results showed that, similar to rats with hippocampal lesions, MEC-lesioned rats were impaired on measures of spatial memory, but not spatial decision-making, with greatest impairments on configurations requiring a global navigational strategy for selecting the optimal route. These findings suggest that the MEC is important for effective spatial navigation, especially when global cue processing is required.