Explore the vast range of titles available.

The increasing deployment of multiple unmanned vehicles systems has generated large research interest in recent decades. This paper therefore provides a detailed survey to review a range of techniques related to the operation of multi-vehicle systems in different environmental domains, including land based, aerospace and marine with the specific focuses placed on formation control and cooperative motion planning. Differing from other related papers, this paper pays a special attention to the collision avoidance problem and specifically discusses and reviews those methods that adopt flexible formation shape to achieve collision avoidance for multi-vehicle systems. In the conclusions, some open research areas with suggested technologies have been proposed to facilitate the future research development.

Unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs) have rapidly evolved, becoming integral to various applications such as environmental monitoring, disaster response, and precision agriculture. This paper provides a comprehensive review of the advancements and the challenges in UAV-UGV collaboration and its potential applications. These systems offer enhanced situational awareness and operational efficiency, enabling complex tasks that are beyond the capabilities of individual systems by leveraging the complementary strengths of UAVs and UGVs. Key areas explored in this review include multi-UAV and multi-UGV systems, collaborative aerial and ground operations, and the communication and coordination mechanisms that support these collaborative efforts. Furthermore, this paper discusses potential limitations, challenges and future research directions, and considers issues such as computational constraints, communication network instability, and environmental adaptability. The review also provides a detailed analysis of how these issues impact the effectiveness of UAV-UGV collaboration.

The use of unmanned aerial vehicles (UAVs) for different applications has increased tremendously during the past decade. The small size, high maneuverability, ability to fly at predetermined coordinates, simple construction, and affordable price have made UAVs a popular choice for diverse aerial applications. However, the small size and the ability to fly close to the terrain make the detection and tracking of UAVs challenging. Similarly, unmanned underwater vehicles (UUVs) have revolutionized underwater operations. UUVs can accomplish numerous tasks that were not possible with manned underwater vehicles. In this survey paper, we provide features and capabilities expected from current and future UAVs and UUVs, and review potential challenges and threats due to use of such UAVs/UUVs. We also overview the countermeasures against such threats, including approaches for the detection, tracking, and classification of UAVs and UUVs.

Ongoing advancements in Remotely Operated Vehicles (ROVs) have enabled their widespread use in various safety-critical applications. However, operators often find themselves controlling the vehicles in harsh and stressful conditions, which can induce stress and fatigue. Such factors may compromise the mission’s safety and outcome, as the operators might issue unintentional commands. Nonetheless, real-time monitoring of both the operator and the ROV can help prevent such potential accidents by introducing a mechanism that detects such anomalies. We present the construction of real-life datasets that include two test cases: aerial and ground vehicles. Data were collected from the operator and the ROV during a mission. The first dataset consists of data from 19 subjects when operating an Unmanned Aerial Vehicle (UAV), while the second dataset includes data from 7 subjects when performing an operation with an Unmanned Ground Vehicle (UGV). The construction of such datasets and the expansion in more than one ROV aim to the generalization of our approach towards abnormal command detection. A thorough analysis was conducted and presented, which included statistical analysis: a t-test and the extraction of average values and box-plots. Further, feature extraction and selection were performed as part of the analysis of the constructed datasets, towards the classification of abnormal commands.

Interest in Unmanned Aerial Vehicles (UAVs) has increased due to their versatility and variety of applications, however their battery life limits their applications. Heterogeneous multi-robot systems can offer a solution to this limitation, by allowing an Unmanned Ground Vehicle (UGV) to serve as a recharging station for the aerial one. Moreover, cooperation between aerial and terrestrial robots allows them to overcome other individual limitations, such as communication link coverage or accessibility, and to solve highly complex tasks, e.g., environment exploration, infrastructure inspection or search and rescue. This work proposes a vision-based approach that enables an aerial robot to autonomously detect, follow, and land on a mobile ground platform. For this purpose, ArUcO fiducial markers are used to estimate the relative pose between the UAV and UGV by processing RGB images provided by a monocular camera on board the UAV. The pose estimation is fed to a trajectory planner and four decoupled controllers to generate speed set-points relative to the UAV. Using a cascade loop strategy, these set-points are then sent to the UAV autopilot for inner loop control. The proposed solution has been tested both in simulation, with a digital twin of a solar farm using ROS, Gazebo and Ardupilot Software-in-the-Loop (SiL); and in the real world at IST Lisbon’s outdoor facilities, with a UAV built on the basis of a DJ550 Hexacopter and a modified Jackal ground robot from DJI and Clearpath Robotics, respectively. Pose estimation, trajectory planning and speed set-point are computed on board the UAV, using a Single Board Computer (SBC) running Ubuntu and ROS, without the need for external infrastructure.

This study establishes the concept and classification system of MUM-T for the operation and development of AI-based complex combat systems. We analyze the core aspects of this system: autonomy, interoperability, and program level. AI MUM-T can improve the survivability of manned systems, expand their operational range, and increase combat effectiveness. We analyze technical challenges and program levels using data from the USA and UK, which are building the AI MUM-T integrated combat system. Currently, MUM-T is at the level of complex operation of a manned platform and an unmanned aerial vehicle platform. In the mid to long term, interoperable communication between heterogeneous platforms such as unmanned ground vehicles, unmanned surface vehicles, and unmanned underwater vehicles is possible. Depending on the level of development of the common architecture and standard protocols for interoperability between AI MUM-T systems, MUM-T can evolve from the “1 to N” concept to various combinations of operating concepts from “N to N.” The difference of this study from existing studies is that the core technologies of the fourth industrial revolution, such as AI, autonomy, and data interoperability, are reflected in the MUM-T system. In addition, an AI-enabled autonomous MUM-T operation and facility classification system was established by reflecting AI and autonomy in the existing unmanned system taxonomy, and the level and program were analyzed taking this into consideration.

The entropy-oriented approach called security- or cybersecurity-informed safety (SIS or CSIS, respectively) is discussed and developed in order to analyse and evaluate the safety and dependability of autonomous transport systems (ATSs) such as unmanned aerial vehicles (UAVs), unmanned maritime vehicles (UMVs), and satellites. This approach allows for extending and integrating the known techniques FMECA (Failure Modes, Effects, and Criticality Analysis) and IMECA (Intrusion MECA), as well as developing the new SISMECA (SIS-based Intrusion Modes, Effects, and Criticality Analysis) technique. The ontology model and templates for SISMECA implementation are suggested. The methodology of safety assessment is based on (i) the application and enhancement of SISMECA considering the particularities of various ATSs and roles of actors (regulators, developers, operators, customers); (ii) the development of a set of scenarios describing the operation of ATS in conditions of cyberattacks and physical influences; (iii) AI contribution to system protection for the analysed domains; (iv) scenario-based development and analysis of user stories related to different cyber-attacks, as well as ways to protect ATSs from them via AI means/platforms; (v) profiling of AI platform requirements by use of characteristics based on AI quality model, risk-based assessment of cyberattack criticality, and efficiency of countermeasures which actors can implement. Examples of the application of SISMECA assessment are presented and discussed.

Although marine litter monitoring has increased over the years, the pollution of coastal waters is still understudied and there is a need for spatial and temporal data. Aerial (UAV) and underwater (ROV) drones have demonstrated their potential as monitoring tools at coastal sites; however, suitable conditions for use and cost-efficiency of the methods still need attention. This study tested UAVs and ROVs for the monitoring of floating, submerged, and seafloor items using artificial plastic plates and assessed the influence of water conditions (water transparency, color, depth, bottom substrate), item characteristics (color and size), and method settings (flight/dive height) on detection accuracy. A cost-efficiency analysis suggests that both UAV and ROV methods lie within the same cost and efficiency category as current on-boat observation and scuba diving methods and shall be considered for further testing in real scenarios for official marine litter monitoring methods.

Autonomous landing on the deck of a boat or an unmanned surface vehicle (USV) is the minimum requirement for increasing the autonomy of water monitoring missions. This paper introduces an end-to-end control technique based on deep reinforcement learning for landing an unmanned aerial vehicle on a visual marker located on the deck of a USV. The solution proposed consists of a hierarchy of Deep Q-Networks (DQNs) used as high-level navigation policies that address the two phases of the flight: the marker detection and the descending manoeuvre. Few technical improvements have been proposed to stabilize the learning process, such as the combination of vanilla and double DQNs, and a partitioned buffer replay. Simulated studies proved the robustness of the proposed algorithm against different perturbations acting on the marine vessel. The performances obtained are comparable with a state-of-the-art method based on template matching.

We present an unsupervised deep learning approach for post-disaster building damage detection that can transfer to different typologies of damage or geographical locations. Previous advances in this direction were limited by insufficient qualitative training data. We propose to use a state-of-the-art Anomaly Detecting Generative Adversarial Network (ADGAN) because it only requires pre-event imagery of buildings in their undamaged state. This approach aids the post-disaster response phase because the model can be developed in the pre-event phase and rapidly deployed in the post-event phase. We used the xBD dataset, containing pre- and post- event satellite imagery of several disaster-types, and a custom made Unmanned Aerial Vehicle (UAV) dataset, containing post-earthquake imagery. Results showed that models trained on UAV-imagery were capable of detecting earthquake-induced damage. The best performing model for European locations obtained a recall, precision and F1-score of 0.59, 0.97 and 0.74, respectively. Models trained on satellite imagery were capable of detecting damage on the condition that the training dataset was void of vegetation and shadows. In this manner, the best performing model for (wild)fire events yielded a recall, precision and F1-score of 0.78, 0.99 and 0.87, respectively. Compared to other supervised and/or multi-epoch approaches, our results are encouraging. Moreover, in addition to image classifications, we show how contextual information can be used to create detailed damage maps without the need of a dedicated multi-task deep learning framework. Finally, we formulate practical guidelines to apply this single-epoch and unsupervised method to real-world applications.