Search

Two image-based sensing methods are merged to mimic human vision in support of airborne detect-and-avoid and counter–unmanned aircraft systems applications. In the proposed sensing system architect...

With natural disasters continuing to become more prevalent in recent years, the need for effective disaster management efforts becomes even more critical. Specifically, flooding is an extremely common natural disaster which can cause significant damage to homes and other property. In this article, we look at an area in Hurley, Virginia which suffered a significant flood event in August 2021. A drone is used to capture aerial imagery of the area and reconstructed to produce 3-dimensional models, Digital Elevation Models, and stitched orthophotos for flood modeling and damage assessment. Pre-flood Digital Elevation Models and available weather data are used to perform simulations of the flood event using HEC-RAS software. These were validated with measured water height values and found to be very accurate. After this validation, simulations are performed using the Digital Elevation Models collected after the flood and we found that a similar rainfall event on the new terrain would cause even worse flooding, with water depths between 29% and 105% higher. These simulations could be used to guide recovery efforts as well as aid response efforts for any future events. Finally, we look at performing semantic segmentation on the collected aerial imagery to assess damage to property from the flood event. While our segmentation of debris needs more work, it has potential to help determine the extent of damage and aid disaster response. Based on our investigation, the combination of techniques presented in this article has significant potential to aid in preparation, response, and recovery efforts for natural disasters. Published version

Accessible, low-cost technologies and tools are needed in the developing world to support community planning, disaster risk assessment, and land tenure. Enterprise-scale geographic information system (GIS) software and high-resolution aerial or satellite imagery are tools which are typically not available to or affordable for resource-limited communities. In this paper, we present a concept of aerial data collection, 3D cadastre modeling, and disaster risk assessment using low-cost drones and adapted open-source software. Computer vision/machine learning methods are used to create a classified 3D cadastre that contextualizes and quantifies potential natural disaster risk to existing or planned infrastructure. Building type and integrity are determined from aerial imagery. Potential flood damage risk to a building is evaluated as a function of three mechanisms: undermining (erosion) of the foundation, hydraulic pressure damage, and building collapse due to water load. Use of Soil and Water Assessment Tool (SWAT) provides water runoff estimates that are improved using classified land features (urban ecology, erosion marks) to improve flow direction estimates. A convolutional neural network (CNN) is trained to find these flood-induced erosion marks from high-resolution drone imagery. A flood damage potential metric scaled by property value estimates results in individual and community property risk assessments. Published version

Autonomous navigation to a specified waypoint is traditionally accomplished with a layered stack of global path planning and local motion planning modules that generate feasible and obstacle-free trajectories. While these modules can be modified to meet task-specific constraints and user preferences, current modification procedures require substantial effort on the part of an expert roboticist with a great deal of technical training. In this paper, we simplify this process by inserting a Machine Learning module between the global path planning and local motion planning modules of an off-the shelf navigation stack. This model can be trained with human demonstrations of the preferred navigation behavior, using a training procedure based on Behavioral Cloning, allowing for an intuitive modification of the navigation policy by non-technical users to suit task-specific constraints. We find that our approach can successfully adapt a robot’s navigation behavior to become more like that of a demonstrator. Moreover, for a fixed amount of demonstration data, we find that the proposed technique compares favorably to recent baselines with respect to both navigation success rate and trajectory similarity to the demonstrator. Published version

Ground operations in urban environments present a considerable risk to dismounted and mobile tactical teams. Special operations, counterterrorism, and other missions require forces to operate in complex, densely occluded urban environments. In these situations, an adversary can often exploit in-depth local knowledge of the 3D structure of the environment to establish optimal cover, maintain concealed ring positions, and channelize the attacking force into exposed ingress routes. Small Unmanned Aerial Systems (UAS) operating at low altitude and indoors could potentially provide real-time situational awareness in these conditions but are constrained by the limited availability of GPS in these environments and by a lack of detailed 3D map data required for precision sensor-based navigation in complex environments. This paper, based on our prior work in1 introduces first responder and militaristic applications for an open- source framework for the rapid processing, exploitation, and dissemination of 3D mapping and navigational data at the tactical edge of the network. The proposed protocol is called 3D Tiles Nav which includes certain navigation cell and visibility metadata in order to improve navigational performance and autonomy in complex environments. The navigation cell data structures organizes the low airspace (below 400ft) in which 3D Tiles NAV data is embedded. The resulting data organization supports efficient data delivery, especially in densely occluded (e.g. low altitude and ground level) use cases. In addition, the resulting metadata fuses information about the visible and navigable structure of the operating environment to provide a unique informational framework for planning and conducting missions in densely occluded, high-threat environments. We demonstrate through experiments the extraction of planar features from point cloud data and illustrate through simulations the reduction network, compute, storage, and sensor requirements made possible using our approach.

This paper introduces a novel protocol for managing low altitude 3D aeronautical chart data to address the unique navigational challenges and collision risks associated with populated urban environments. Based on the Open Geospatial Consortium (OGC) 3D Tiles standard for geospatial data delivery, the proposed extension, called 3D Tiles Nav., uses a navigation-centric packet structure which automatically decomposes the navigable regions of space into hyperlocal navigation cells and encodes environmental surfaces that are potentially visible from each cell. The developed method is sensor agnostic and provides the ability to quickly and conservatively encode visibility directly from a region by enabling an expanded approach to viewshed analysis. In this approach, the navigation cells themselves are used to represent the intrinsic positional uncertainty often needed for navigation. Furthermore, we present in detail this new data format and its unique features as well as a candidate framework illustrating how an Unmanned Traffic Management (UTM) system could support trajectory-based operations and performance-based navigation in the urban canyon. Our results, experiments, and simulations conclude that this data reorganization enables 3D map streaming using less bandwidth and efficient 3D map-matching systems with limited on-board compute, storage, and sensor resources. Published version

This paper proposes a combined observer/controller method that estimates the states and simultaneously improves the ride comfort and stability of a full vehicle active suspension system using a sin...

We present an approach to simultaneously perform semantic segmentation and prepositional phrase attachment resolution for captioned images. Some ambiguities in language cannot be resolved without simultaneously reasoning about an associated image. If we consider the sentence “I shot an elephant in my pajamas”, looking at language alone (and not using common sense), it is unclear if it is the person or the elephant wearing the pajamas or both. Our approach produces a diverse set of plausible hypotheses for both semantic segmentation and prepositional phrase attachment resolution that are then jointly re-ranked to select the most consistent pair. We show that our semantic segmentation and prepositional phrase attachment resolution modules have complementary strengths, and that joint reasoning produces more accurate results than any module operating in isolation. Multiple hypotheses are also shown to be crucial to improved multiple-module reasoning. Our vision and language approach significantly outperforms the Stanford Parser (De Marneffe et al., 2006) by 17.91% (28.69% relative) and 12.83% (25.28% relative) in two different experiments. We also make small improvements over DeepLab-CRF (Chen et al., 2015).

A novel morphing control surface design employing piezoelectric macro-fiber composite actuators is compared to a servo-actuated system. The comprehensive comparison including aerodynamics, size, weight, power, bandwidth, and reliability has been extended to include flight test comparisons. Three flight vehicles were designed, built, and evaluated: a servo-controlled generic micro-aerial vehicle aircraft and two conformal actuator controlled versions based on thick and thin wing designs. Flight agility and control response of the morphing-actuated and servo-actuated configurations were quantified through state measurement during identical automated maneuvers. The morphing actuation scheme demonstrated control bandwidth that was an order of magnitude greater than for the servo-actuated system, but showed a 12% decrease in roll rate when compared to the servo-actuated baseline aircraft. The flight vehicles allowed system-level comparisons of conventional and morphing control, where conformal actuation occupied less volume, consumed equivalent power as micro servos and provided effective control power for maneuvering.

Unmanned vehicles, equipped with radiation detection sensors, can serve as a valuable aid to personnel responding to radiological incidents. The use of tele-operated ground vehicles avoids human exposure to hazardous environments, which in addition to radioactive contamination, might present other risks to personnel. Autonomous unmanned vehicles using algorithms for radioisotope classification, source localization, and efficient exploration allow these vehicles to conduct surveys with reduced human supervision allowing teams to address larger areas in less time. This work presents systems for autonomous radiation search with results presented in several proof-of-concept demonstrations.

An active suspension based on Linear Quadratic Gaussian (LQG) optimal controller is an effective system for enhancing the ride comfort and handling characteristics of a vehicle. LQG requires a good plant model for success, and this may be difficult to extract using a single inertial measurement device in the presence of noise. This paper presents a method for estimating the vehicle states by measuring both the vehicle bounce and pitch accelerations using an Inertial Measurement Unit (IMU) with position uncertainty relative to the sprung mass center of gravity. Frequency domain methods are used for System Identification (SysId). The state estimation is based on channel-by-channel model estimation using uncorrelated random excitation which is applied to the front wheels, rear wheels, front actuator, and rear actuator. An anti-aliasing filter eliminates false response harmonics and a Kalman filter is used to estimate the current states of the actual plant and the LQR block for the full-states-feedback controller. The controllers and observer are implemented in simulation using a four degree-of-freedom half car linear model.

As rocks are transported along the conveyor belt of a quarry, the maximum dimension of the rocks exiting the crushers should not exceed a size threshold specific to each crusher. If the rocks are too large then they can pose a threat to equipment, and lead to a large cost in repair and loss of production. A 2D vision system is presented, which is capable of estimating the size distribution of the rocks, and also monitoring for oversize rocks. Image segmentation is performed, which is followed by a process that classifies the segments as valid or invalid using a support vector machine. A novel split algorithm is presented, which attempts to split segments that have resulted in undersegmentation. This allows the system to constantly monitor for oversize rocks without stopping the conveyor belt. For the experiments presented in this paper, a set of images was taken of rocks on a moving conveyor. In testing, it was found that 81.31 % of the segments output by the system correctly found the maximum dimension of the rock that it represented.

An unmanned remote sensing and sampling system has been developed to aid first responders in urban disaster assessment and recovery. The system design is based on a 90 kg autonomous helicopter platform with interchangeable payloads ranging from radiation detection to tethered robot deployment. A typical response would begin with three-dimensional terrain mapping using the stereovision system and a survey of radiation levels with an onboard spectrometer. From this initial survey, a more targeted flight is planned at a lower altitude with the option to localize radioactive sources in the event that radiation is present. Finally, a robot can be tether-deployed into the area of interest to collect samples. It is teleoperated from the ground control station, and after collection is finished the robot is retracted back to the helicopter for retrieval. The terrain mapping, radiation detection, radiation localization, and robot deployment and retrieval have all been flight-tested. Results of these tests indicate that the systems functioned successfully in the context of a prototype demonstrator.

Engineering efforts in low resource environments pose a unique set of challenges, requiring an in-depth understanding of local needs, comprehensive mapping of community resources, and extensive collaboration with local expertise. The importance of these principles is demonstrated in this paper by detailing the novel design and field demonstration of an affordable, locally manufactured intravenous fluid regulation device. Collaboration with clinical personnel in Uganda and Malawi guided device design. In-country physicians emphasised the need to regulate volume of intravenous (IV) fluid delivered to a paediatric patient without use of electricity. The proposed device regulates IV fluid delivery within ±20 mL of total prescribed dosage, providing a method of reducing fatalities caused by over-hydration in low resource environments; the feasibility of building the device from local resources was demonstrated by a field research team in Malawi. The device was successfully constructed entirely from local resources for a total cost of $46.21 (USD). Additionally, the device was demonstrated in rural clinics where 89 % of surveyed clinical staff reported that they would use the device to regulate IV fluid delivery. This paper emphasises the importance of collaborating with communities for community-based engineering solutions. Mapping community assets and collaborating with local expertise are crucial to success of engineering efforts. Long-term, community-based efforts are likely to sustainably improve health outcomes and strengthen economies of communities worldwide.

Recent events have highlighted the need for unmanned remote sensing in dangerous areas, particularly where structures have collapsed or explosions have occurred, to limit hazards to first responders and increase their efficiency in planning response operations. In the case of the Fukushima nuclear reactor explosion, an unmanned helicopter capable of obtaining overhead images, gathering radiation measurements, and mapping both the structural and radiation content of the environment would have given the response team invaluable data early in the disaster, thereby allowing them to understand the extent of the damage and areas where dangers to personnel existed. With this motivation, the Unmanned Systems Lab at Virginia Tech has developed a remote sensing system for radiation detection and aerial imaging using a 90 kg autonomous helicopter and sensing payloads for the radiation detection and imaging operations. The radiation payload, which is the sensor of focus in this paper, consists of a scintillating type detector with associated software and novel search algorithms to rapidly and effectively map and locate sources of high radiation intensity. By incorporating this sensing technology into an unmanned aerial vehicle system, crucial situational awareness can be gathered about a post-disaster environment and response efforts can be expedited. This paper details the radiation mapping and localization capabilities of this system as well as the testing of the various search algorithms using simulated radiation data. The various components of the system have been flight tested over a several-year period and a new production flight platform has been built to enhance reliability and maintainability. The new system is based on the Aeroscout B1-100 helicopter platform, which has a one-hour flight endurance and uses a COFDM radio system that gives the helicopter an effective range of 7 km.

Aerial terrain mapping has been used for many years to monitor natural habitats and ecosystems, assist in urban planning, and monitor trends in land usage. Recent improvements in digital imaging, L...

The purpose here is to enable peak-to-peak actuation of piezocomposite devices to perform shape control of structures, specifically bimorph variable-camber control surfaces on small aircraft. One of the major drawbacks in implementing control with piezocomposite actuation is its asymmetry in the positive and negative voltage directions. To remedy this situation, a novel, solid-state electrical circuit employing diodes and resistors is proposed. The circuit allows a single bipolar amplifier to control a bimorph composed of two piezoceramic composite actuators known as Macro-Fiber Composite (MFC). The MFCs have an asymmetric actuation range of -500 to 1500 V; hence an MFC bimorph with conventional serial or parallel electrical connection limits the actuation to the -500 to 500 V range. The proposed circuit allows the division of the input voltage so that: (1) The MFC that is in extension receives 1500 V, and (2) The MFC that is in compression receives -500 V. The circuit also allows the actuation in the opposite direction without any physical changes (i.e., not using any electromechanical switches) The input-output relationship of the circuit is characterized experimentally. An MFC bimorph is used to further demonstrate the capability of the circuit.

The research presented here employs solid-state actuators for flow separation delay or for forced attachment of separated flow seen in airfoils at low Reynolds numbers. To reduce separation, periodic excitation to the flow around the leading edge of the airfoil is induced by Macro-Fiber Composite actuated clamped-free unimorph benders. An electromechanical model of the unimorph is briefly presented and parametric study is conducted to aid the design of a unimorph to output high deformation at a desired frequency. The optimum frequency and amplitude for lift improvement at post-stall angles are identified experimentally. Along with aerodynamic force and structural displacement measurements, helium bubble flow visualization is used to verify existing separated flow, and the attached flow induced by flow control. The lift enhancement induced by several flow control techniques is compared. A symmetric and non-uniform (3D) flow excitation results in the maximum lift enhancement at post-stall region at the lowest power consumption level. A maximum lift coefficient increase of 27.5% (in the post-stall region) is achieved at 125 Hz periodic excitation, with the 3D symmetric actuation mode at 5 m/s and the reduced frequency of 3.78. Cl, max is increased 7.6% from the baseline.

This study aims to enable solid-state aerodynamic force generation in high-dynamic-pressure airflow. A novel, high-load-output, bidirectional variable-camber airfoil employing a type of piezoceramic composite actuator known as a Macro-Fiber Composite is presented. The novel airfoil employs two active surfaces and a single four-bar (box) mechanism as the internal structure. The unique choice of boundary conditions allows variable and smooth deformation in both directions from a flat camber line. The paper focuses on actuation modeling and response characterization under aerodynamic loads. A parametric study of aerodynamic response is employed to optimize the kinematic parameters of the airfoil. The concept is fabricated by implementing eight Macro-Fiber Composite 8557-P1-type actuators in a bimorph configuration to construct the active surfaces. The box mechanism generates deflection and camber change as predicted. Wind-tunnel experiments are conducted on a 12.6% maximum thickness, 127 mm chord airfoil. Aerodynamic and structural performance results are presented for a flow rate of 15 m /s and a Reynolds number of 127,000. Nonlinear effects due to aerodynamic and piezoceramic hysteresis are identified and discussed. A lift coefficient change of 1.54 is observed purely due to voltage actuation. Results are compared with conventional, zero-camber NACA and other airfoils. A 72% increase in the lift-curve slope is achieved when compared with a NACA 0009 airfoil.