Unmanned Ground Vehicle Operating Research Articles

Aerial online mapping on‐board system by real‐time object detection for UGV path generation in unstructured outdoor environments

AbstractAn optimal path provides efficient operation of unmanned ground vehicles (UGVs) for many kinds of tasks such as transportation, exploration, surveillance, and search and rescue in unstructured areas that include various unexpected obstacles. Various onboard sensors such as LiDAR, radar, sonar, and cameras are used to detect obstacles around the UGVs. However, their range of view is often limited by movable obstacles or barriers, resulting in inefficient path generation. Here, we present the aerial online mapping system to generate an efficient path for a UGV on a two‐dimensional map. The map is updated by projecting obstacles detected in the aerial images taken by an unmanned aerial vehicle through an object detector based on a conventional convolutional neural network. The proposed system is implemented in real‐time by a skid steering ground vehicle and a quadcopter with relatively small, low‐cost embedded systems. The frameworks and each module of the systems are given in detail to evaluate the performance. The system is also demonstrated in unstructured outdoor environments such as in a football field and a park with unreliable communication links. The results show that the aerial online mapping is effective in path generation for autonomous UGVs in real environments.

In-Device Feedback in Immersive Head-Mounted Displays for Distance Perception During Teleoperation of Unmanned Ground Vehicles.

In recent years, Virtual Reality (VR) Head-Mounted Displays (HMD) have been used to provide an immersive, first-person view in real-time for the remote-control of Unmanned Ground Vehicles (UGV). One critical issue is that it is challenging to perceive the distance of obstacles surrounding the vehicle from 2D views in the HMD, which deteriorates the control of UGV. Conventional distance indicators used in HMD take up screen space which leads clutter on the display and can further reduce situation awareness of the physical environment. To address the issue, in this paper we propose off-screen in-device feedback using vibro-tactile and/or light-visual cues to provide real-time distance information for the remote control of UGV. Results from a study show a significantly better performance with either feedback type, reduced workload and improved usability in a driving task that requires continuous perception of the distance between the UGV and its environmental objects or obstacles. Our findings show a solid case for in-device vibro-tactile and/or light-visual feedback to support remote operation of UGVs that highly relies on distance perception of objects.

Research on a multi-attribute comprehensive evaluation system for unmanned ground vehicles

To promote the development of unmanned ground vehicles, it is very necessary to establish a scientific and reasonable evaluation system for unmanned ground vehicles. This paper analyses the structure and operation of unmanned ground vehicles, and also proposes a multi-attribute comprehensive evaluation system for unmanned ground vehicles. Based on six aspects including environment perception, communication command, behavioural decision-making, human-computer interaction, weapon system and mobile platform, a new evaluation indicator system is established. In the quantitative process of qualitative indicators, Z-number theory is introduced to better describe the uncertainty degree of judgment and reasoning in human’s understanding of complex things. In order to make the indicator weights more scientific, the combined weighting model with analytic hierarchy process and information entropy is adopted for weight analysis, which not only takes into account the influence of evaluation experts on weights, but also reduces the subjective arbitrariness of weights. The least-squares grey relational analysis method is used to analyse the correlation between the comparison series and the reference series, and a comprehensive quantitative evaluation result of the unmanned ground vehicles is obtained. The evaluation examples of unmanned ground vehicles show that the proposed evaluation system can quantitatively evaluate the overall technical performance of unmanned ground vehicles. This evaluation system has high reference value for the evaluation of other complex weapons.

Sensor-based, time-critical mobility of autonomous robots in cluttered spaces: a harmonic potential approach

SUMMARYThis paper suggests an integrated navigation system for an unmanned ground vehicle operating in an unknown cluttered environment. The navigator supports time-critical mobility making it possible for a mobile robot to reach a target from the first attempt without the need for a dedicated exploration and mapping stage. The robot only uses necessary and sufficient egocentric local sensory data collected on its way to the target. The construction of the navigation structure revolves around key properties of the harmonic potential field approach to motion planning. The robot's trajectory is well-behaved and direct-to-the-goal. It contains only the minimum number of detours necessary to accommodate the sensory data and maintain the robot in a safe, goal-oriented state. The navigation structure is developed and its theoretical basis is explained. Extensive experimental validation of its properties and performance is carried-out using the X80 robotic platform.

Motion control design for unmanned ground vehicle in dynamic environment using intelligent controller

PurposeThe motion control of unmanned ground vehicles (UGV) is a challenge in the industry of automation. The purpose of this paper is to propose a fuzzy inference system (FIS) based on sensory information for solving the navigation challenge of UGV in cluttered and dynamic environments.Design/methodology/approachThe representation of the dynamic environment is a key element for the operational field and for the testing of the robotic navigation system. If dynamic obstacles move randomly in the operation field, the navigation problem becomes more complicated due to the coordination of the elements for accurate navigation and collision-free path within the environmental representations. This paper considers the construction of the FIS, which consists of two controllers. The first controller uses three sensors based on the obstacles distances from the front, right and left. The second controller employs the angle difference between the heading of the vehicle and the targeted angle to obtain the optimal route based on the environment and reach the desired destination with minimal running power and delay. The proposed design shows an efficient navigation strategy that overcomes the current navigation challenges in dynamic environments.FindingsExperimental analyses are conducted for three different scenarios to investigate the validation and effectiveness of the introduced controllers based on the FIS. The reported simulation results are obtained using MATLAB software package. The results show that the controllers of the FIS consistently perform the manoeuvring task and manage the route plan efficiently, even in a complex environment that is populated with dynamic obstacles. The paper demonstrates that the destination was reached optimally using the shortest free route.Research limitations/implicationsThe paper represents efforts toward building a dynamic environment filled with dynamic obstacles that move at various speeds and directions. The methodology of designing the FIS is accomplished to guide the UGV to the desired destination while avoiding collisions with obstacles. However, the methodology is approached using two-dimensional analyses. Hence, the paper suggests several extensions and variations to develop a three-dimensional strategy for further improvement.Originality/valueThis paper presents the design of a FIS and its characterizations in dynamic environments, specifically for obstacles that move at different velocities. This facilitates an improved functionality of the operation of UGV.

A Green Ant-Based method for Path Planning of Unmanned Ground Vehicles

Planning of optimal/shortest path is required for proper operation of unmanned ground vehicles (UGVs). Although most of the existing approaches provide proper path planning strategy, they cannot guarantee reduction of consumed energy by UGVs, which is provided via onboard battery with constraint power. Hence, in this paper, a new ant-based path planning approach that considers UGV energy consumption in its planning strategy is proposed. This method is called Green Ant (G-Ant) and integrates an ant-based algorithm with a power/energy consumption prediction model to reach its main goal, which is providing a collision-free shortest path with low power consumption. G-Ant is evaluated and validated via simulation tools. Its performance is compared with ant colony optimization, genetic algorithm, and particle swarm optimization approaches. Various scenarios were simulated to evaluate G-Ant performance in terms of UGV travel time, travel length, computational time by taking into account different numbers of iterations, different numbers of obstacle, and different population sizes. The obtained results show that the G-Ant outperforms the existing methods in terms of travel length and number of iteration.

A Self‐learning Framework for Statistical Ground Classification using Radar and Monocular Vision

Reliable terrain analysis is a key requirement for a mobile robot to operate safely in challenging environments, such as in natural outdoor settings. In these contexts, conventional navigation systems that assume a priori knowledge of the terrain geometric properties, appearance properties, or both, would most likely fail, due to the high variability of the terrain characteristics and environmental conditions. In this paper, a self‐learning framework for ground detection and classification is introduced, where the terrain model is automatically initialized at the beginning of the vehicle's operation and progressively updated online. The proposed approach is of general applicability for a robot's perception purposes, and it can be implemented using a single sensor or combining different sensor modalities. In the context of this paper, two ground classification modules are presented: one based on radar data, and one based on monocular vision and supervised by the radar classifier. Both of them rely on online learning strategies to build a statistical feature‐based model of the ground, and both implement a Mahalanobis distance classification approach for ground segmentation in their respective fields of view. In detail, the radar classifier analyzes radar observations to obtain an estimate of the ground surface location based on a set of radar features. The output of the radar classifier serves as well to provide training labels to the visual classification module. Once trained, the vision‐based classifier is able to discriminate between ground and nonground regions in the entire field of view of the camera. It can also detect multiple terrain components within the broad ground class. Experimental results, obtained with an unmanned ground vehicle operating in a rural environment, are presented to validate the system. It is shown that the proposed approach is effective in detecting drivable surface, reaching an average classification accuracy of about 80% on the entire video frame with the additional advantage of not requiring human intervention for training or a priori assumption on the ground appearance.

Cornering Law

We investigated skill development and workload of pilots driving teleoperated unmanned ground vehicles (UGVs) through different apertures and viewpoints using the cornering law. Due to technological and cost restraints, humans are still needed for tasks involving UGVs. Operators of teleoperated UGVs are likely to have less situation awareness and thus are more prone to getting stuck or damaged when negotiating apertures.To our knowledge, the operation of physical UGVs through corners has not been examined. Therefore, a better understanding of cornering a teleoperated UGVs is imperative. In Experiment 1, 20 novice participants repeatedly teleoperated a physical UGV using a third-person overhead view through apertures that varied in width. In Experiment 2, 18 additional novice participants completed a similar task but used a first-person view. Participants' performance increased (i.e., faster cornering times and less collisions) over sessions. The cornering law successfully modeled the effect of different aperture widths on participant performance for both viewing perspectives. In this study, we successfully modeled human performance of teleoperated UGVs using the cornering law. Analogous to Fitts' and steering law, we were able to successfully model and predict cornering performance based on a derived index of cornering difficulty. The cornering law could be used to aid in the development of prototype user interfaces and also to examine the effects of different teleoperation views (first person vs. third person).

Self-learning classification of radar features for scene understanding

Autonomous driving is a challenging problem in mobile robotics, particularly when the domain is unstructured, as in an outdoor setting. In addition, field scenarios are often characterized by low visibility as well, due to changes in lighting conditions, weather phenomena including fog, rain, snow and hail, or the presence of dust clouds and smoke. Thus, advanced perception systems are primarily required for an off-road robot to sense and understand its environment recognizing artificial and natural structures, topology, vegetation and paths, while ensuring, at the same time, robustness under compromised visibility. In this paper the use of millimeter-wave radar is proposed as a possible solution for all-weather off-road perception. A self-learning approach is developed to train a classifier for radar image interpretation and autonomous navigation. The proposed classifier features two main stages: an adaptive training stage and a classification stage. During the training stage, the system automatically learns to associate the appearance of radar data with class labels. Then, it makes predictions based on past observations. The training set is continuously updated online using the latest radar readings, thus making it feasible to use the system for long range and long duration navigation, over changing environments. Experimental results, obtained with an unmanned ground vehicle operating in a rural environment, are presented to validate this approach. A quantitative comparison with laser data is also included showing good range accuracy and mapping ability as well. Finally, conclusions are drawn on the utility of millimeter-wave radar as a robotic sensor for persistent and accurate perception in natural scenarios.

Some Good and Bad with Spatial Ability in Three Person Teams That Operate Multiple Unmanned Vehicles

This study reports findings regarding the influence of spatial ability of each operator on a three person team on workload and performance. Sixty six participants were randomly assigned to the role of unmanned aerial vehicle (UAV) operator, unmanned ground vehicle (UGV) operator, and intelligence officer (leader) to create a total of 22 teams, and spatial ability was assessed with Part 5 of the Guilford-Zimmerman Aptitude Survey. Findings indicated that spatial ability of the UAV operator and UGV operator improved reconnaissance, and while spatial ability of the UAV operator improved reacquisition of objectives after reconnaissance, spatial ability of the intelligence officer hindered team performance on this second task. A rationale for these results was developed with findings from the Multiple Resource Questionnaire (MRQ). Discussion focuses on the relationship between spatial ability and visual perception in complex teams.

Interactive Effects of Backup Behavior and Spatial Abilities in the Prediction of Teammate Workload Using Multiple Unmanned Vehicles

This study examined the interactive effects of spatial ability and team process on operator workload, while using multiple unmanned vehicles. The hypotheses also focused on how these effects might change when using different measures of spatial ability. In order to examine this, the Guilford-Zimmerman Spatial Visualization and Spatial Orientation scores of an unmanned aerial vehicle (UAV) operator and navigation support provided by this UAV operator to an unmanned ground vehicle (UGV) operator were used as variables predicting UGV operator workload while performing a reconnaissance task. Results indicated that the interaction of the “guider's” spatial visualization and navigation support and the interaction between spatial orientation and navigation support not only accounted for unique variances in the prediction of his/her teammate's workload, but they also produced qualitatively different patterns of results. In identifying these unique contributions, the importance of using multiple spatial ability measures in (unmanned vehicle) research is highlighted.

Considerations for Use of Aerial Views in Remote Unmanned Ground Vehicle Operations

Remote unmanned ground vehicle (UGV) operations place the human operator at a perceptual disadvantage. Adding aerial views can benefit the operator's spatial cognition by supplying the missing contextual information regarding the vehicle's position and relation to other objects in the space surrounding the vehicle. In order to benefit from this additional information the operator must control and integrate multiple viewpoints. In a series of experiments the use of aerial views was examined including control mode options and altitude for the aerial scene imaging. Results indicate that aerial views are beneficial in UGV search tasks and that auto-tracking aerial imaging control modes should be considered.

Visual-cognition challenges for the operation of unmanned ground vehicles

Visual-cognition challenges for the operation of unmanned ground vehicles