Search

Coral reefs are among the most diverse ecosystems on our planet, and are depended on by hundreds of millions of people. Unfortunately, most coral reefs are existentially threatened by global climate change and local anthropogenic pressures. To better understand the dynamics underlying deterioration of reefs, monitoring at high spatial and temporal resolution is key. However, conventional monitoring methods for quantifying coral cover and species abundance are limited in scale due to the extensive manual labor required. Although computer vision tools have been employed to aid in this process, in particular SfM photogrammetry for 3D mapping and deep neural networks for image segmentation, analysis of the data products creates a bottleneck, effectively limiting their scalability. This paper presents a new paradigm for mapping underwater environments from ego-motion video, unifying 3D mapping systems that use machine learning to adapt to challenging conditions under water, combined with a modern approach for semantic segmentation of images. The method is exemplified on coral reefs in the northern Gulf of Aqaba, Red Sea, demonstrating high-precision 3D semantic mapping at unprecedented scale with significantly reduced required labor costs: a 100 m video transect acquired within 5 minutes of diving with a cheap consumer-grade camera can be fully automatically analyzed within 5 minutes. Our approach significantly scales up coral reef monitoring by taking a leap towards fully automatic analysis of video transects. The method democratizes coral reef transects by reducing the labor, equipment, logistics, and computing cost. This can help to inform conservation policies more efficiently. The underlying computational method of learning-based Structure-from-Motion has broad implications for fast low-cost mapping of underwater environments other than coral reefs.

Countless signal processing applications include the reconstruction of signals from few indirect linear measurements. The design of effective measurement operators is typically constrained by the underlying hardware and physics, posing a challenging and often even discrete optimization task. While the potential of gradient-based learning via the unrolling of iterative recovery algorithms has been demonstrated, it has remained unclear how to leverage this technique when the set of admissible measurement operators is structured and discrete. We tackle this problem by combining unrolled optimization with Gumbel reparametrizations, which enable the computation of low-variance gradient estimates of categorical random variables. Our approach is formalized by GLODISMO (Gradient-based Learning of DIscrete Structured Measurement Operators). This novel method is easy-to-implement, computationally efficient, and extendable due to its compatibility with automatic differentiation. We empirically demonstrate the performance and flexibility of GLODISMO in several prototypical signal recovery applications, verifying that the learned measurement matrices outperform conventional designs based on randomization as well as discrete optimization baselines., Comment: The current version of the manuscript has been accepted to IEEE Journal on Selected Areas in Information Theory

It is well-established that many iterative sparse reconstruction algorithms can be unrolled to yield a learnable neural network for improved empirical performance. A prime example is learned ISTA (LISTA) where weights, step sizes and thresholds are learned from training data. Recently, Analytic LISTA (ALISTA) has been introduced, combining the strong empirical performance of a fully learned approach like LISTA, while retaining theoretical guarantees of classical compressed sensing algorithms and significantly reducing the number of parameters to learn. However, these parameters are trained to work in expectation, often leading to suboptimal reconstruction of individual targets. In this work we therefore introduce Neurally Augmented ALISTA, in which an LSTM network is used to compute step sizes and thresholds individually for each target vector during reconstruction. This adaptive approach is theoretically motivated by revisiting the recovery guarantees of ALISTA. We show that our approach further improves empirical performance in sparse reconstruction, in particular outperforming existing algorithms by an increasing margin as the compression ratio becomes more challenging., Comment: 10pages, 9 figures

Point clouds provide a flexible and natural representation usable in countless applications such as robotics or self-driving cars. Recently, deep neural networks operating on raw point cloud data have shown promising results on supervised learning tasks such as object classification and semantic segmentation. While massive point cloud datasets can be captured using modern scanning technology, manually labelling such large 3D point clouds for supervised learning tasks is a cumbersome process. This necessitates methods that can learn from unlabelled data to significantly reduce the number of annotated samples needed in supervised learning. We propose a self-supervised learning task for deep learning on raw point cloud data in which a neural network is trained to reconstruct point clouds whose parts have been randomly rearranged. While solving this task, representations that capture semantic properties of the point cloud are learned. Our method is agnostic of network architecture and outperforms current unsupervised learning approaches in downstream object classification tasks. We show experimentally, that pre-training with our method before supervised training improves the performance of state-of-the-art models and significantly improves sample efficiency.

This paper presents a novel flexure design for small scale refelectarrays, specifically CubeSats, that requires high level of deployment precision. Advantages of using a HSC flexure includes additional thermal stability, lesser part count and passive deployment mechanism using the stored strain energy. The proposed design is to use two flat HSC strips of thickness ℎ, length 2𝐿ℎ, width 𝑊ℎ that attach to the core-reflector panel interface. The flexure satisfies two main requirements: ability to fold to a specific compaction 𝑑 without material failure and sufficient stiffness in the deployed configuration. The elastica analysis was implemented to identify the maximum strains in the flexure when folded and provide relations that combine Δ, ℎ, 𝐿ℎ with the maximum allowable strain. A simplified analysis was performed of the vibration of a panel-flexure combination using a two degree of freedom system. The vibration analysis reveals a dimensionless length parameter 𝛽 = (𝐿ℎ/𝐿𝑃) (Δ/ℎ) that determines the transition between bending and shear modes. A non-dimensional relationship was obtained between the natural frequency and 𝛽 that captures the impact of each parameter on the response of the structure. Our current work involves fabricating and testing of the proposed flexures using High Strain Composite materials, with plans to build a prototype for a deployable reflectarray.

Herein is presented a design for a lightweight 5m 1m radio-frequency (RF) re ector that can be attened, folded, and coiled for compact stowage on board an ESPA-class spacecraft. The re ector is a re ectarray that operates at 3.2GHz. Compared to state- of-the-art technologies for RF re ectors, the presented design has advantages in terms of areal density, sti ness, deployed stability, and scalability. Thermal and structural analysis is presented to demonstrate deployed sti ness and thermoelastic stability of the proposed design. Thermal analysis is used to predict in- space deployed temperatures in an operational condition, and structural nite element analysis is used to predict deployed vibration modes and frequencies, and the thermoelastic deformation of the deployed re ector. Also presented are the fabrication, assembly, and testing of two one-third-scale-length full-scale-width 1:7m 1m test articles. These test articles are used to experimentally demonstrate RF functioning, stowage, deployment, and RF performance after deployment.

The Active Thermal Architecture (ATA) is an advanced sub-1U Active Thermal Control technology (ATC) for high power payload support in 6U CubeSat form factors and above. The design utilizes a two-stage, single-phase mechanically pumped fluid loop coupled through a two-axis flexible rotary fluid hinge, to reject thermal power to a deployable tracking radiator. A COTS Ricor K508N cryocooler forms the second stage and provides cryogenic cooling to a custom Kevlar detector mount through a TMT pyrolytic graphene thermal strap. Passive vibration isolation and damping technologies prevent the transfer of jitter to the satellite systems. The ATA design utilizes state-of-the-art 3D fabrication techniques such as Ultrasonic Additive Manufacturing (UAM) to directly embed the working fluid channels into the HX, radiator, and CubeSat chassis allowing for the miniaturization and simplification of the ATA system into an integrated thermal control solution. This paper will focus on the design and ground-based characterization and qualification of the ATA system and provide performance metrics for its use as a thermal support subsystem for advanced infrared electro-optical CubeSat payloads. The ATA project is funded through a NASA Small Satellite Technology Program (SSTP) and is a partnership between the Center for Space Engineering at Utah State University and the Jet Propulsions Laboratory. The ATA active thermal control system has been raised to a TRL of 6 and hopes to provide payload support to advanced missions such as the SABER-Lite and JPL CIRAS projects.

In the past decade CubeSats have made their way into the spotlight. They have evolved from small, university educational opportunities, to industry and governments using them make new discoveries and monetize space. However, with the small, constrained CubeSat form factor; there is often a need to expand the CubeSat through deployable mechanisms once the satellite is in space. This paper is a survey of deployable structures and their actuating mechanisms for CubeSats. The goal of this paper is to provide the applications within which deployable structures have been used in the past for CubeSats, the mechanisms with regards to how they deploy, the lessons learned, and limitations of the various types of deployables. The inputs to this paper come from a relational database in development to track launched CubeSat missions with deployable structures. From this database we can find insightful trends. This paper specifically focuses on the first decade of CubeSat deployables, from 2000 to 2010.

UNKNOWN

The interstellar medium (ISM) represents the next frontier in space exploration, with many new discoveries to be made. The challenge, being so far away from Earth, the ISM requires many decades to reach. To advance our knowledge of what exists beyond our Solar System, new approaches for rapid access are required. One such approach is solar thermal propulsion (STP). The approach uses several Venus and Earth gravity assists to fly to Jupiter and use its gravity well to dive towards the Sun. Approaching within 3 solar radii a perihelion burn would be performed, maximizing the spacecraft’s ΔV to achieve high Solar System escape velocities. A unique aspect of the STP mission concept is that the Sun is not only used as a gravity well for an Oberth manoeuvre, but also to heat the fuel to ultra-high-temperatures (>3000 K), enabling a monopropellant burn with high specific impulse (Isp). Prior preliminary studies indicated escape velocities of over 20 astronomical unit (AU)/yr would be possible. An in-depth modelling exercise was undertaken to determine how such a system would perform. The model in this paper showed the current STP design is capable of providing just under 9±1 AU/yr, but there are many technology developments that could increase escape velocity. The technologies vary from items that could be implemented in the near term, like turbo-pumps driven by the hydrogen, to items requiring more extensive development programs like thin coatings which do not erode in superheated hydrogen. After reviewing the STP approach, and comparing it to a solid rocket motor (SRM), it was found that with currently available technology, SRM outperforms STP with an escape velocity of approximately 10-12 AU/yr. However, future advances in heat exchanger lining materials, turbo pumps, and advanced heat exchanger geometries may enable solar thermal propulsion to provide higher escape velocities, providing one of the fastest ways to exit the solar system. Ultimately, if all technology paths could be implemented with minimal side effects, the performance in a best-case scenario could reach up to 16 AU/yr.

In the past decade, CubeSats have undergone a revolution, moving from universityresearch projects to enabling industry opportunities and government missions. Six yearsago, the Jet Propulsion Laboratory, California Institute of Technology (JPL/Caltech)initiated a research and technology development effort to advance CubeSat communicationcapabilities, with one of the key thrusts being the Ka-band parabolic deployable antenna(KaPDA). This antenna started with the ambitious goal of fitting a 42 dB, 0.5 meter, 35 GHzantenna in a 1.5U canister. At that time, there had been very limited development in the areaof high gain CubeSat antennas which are critical for both high data rate communicationsand remote sensing science. A Ka-band high gain antenna would provide a 10,000 timesincrease in data communication rates over an X-band patch antenna and a 100 timesincrease over state-of-the-art S-band parabolic antennas. This paper discusses the process ofbuilding, integrating, and operating the flight antenna, its final performance and lessonslearned. KaPDA was an enabling technology for RainCube mission, the first Earth ScienceCubeSat to have an active instrument. RainCube was launched in May of 2018, makingKaPDA the second deployable parabolic antenna to fly on a CubeSat and the first of its kindto operate at Ka-band enabling a number of opportunities for high rate, deep space antennacommunications and remote sensing science.

UNKNOWN

UNKNOWN

UNKNOWN