Articles in 2022

Tong Ye and colleagues use a spectrum analyzer and the probability-maintained noise power ratio method to estimate the performance of nonlinear communications systems. The method is verified in seven different nonlinear mechanism and application scenarios.

Izzo and Gómez present a machine learning-based method for obtaining accurate density models of even irregular celestial bodies using minimal prior information. The work is validated on uniform and non-uniform density models of several visited asteroids.

Shi and colleagues create a miniature curvature sensor by integrating a light emitting diode and GaN-based photodetector into a plastic optical fiber. They detect finger motion to show potential for wearable sensing systems.

Asim Waqas and colleagues investigated the relationship between the architectures of deep artificial neural networks (DANNs) and their robustness to noise and adversarial attacks in computer vision. The researchers found that the robustness of DANNs was highly correlated with graph-theoretic measures of entropy and curvature. This finding could help design more robust DANN architectures.

Renn and Gharib experimentally investigate the application of reinforcement learning to provide integrated flow information for aerodynamic control of a wing system in a highly turbulent environment. The results can inform future gust mitigation systems for unmanned aerial vehicles and wind turbines.

Kaitlin Stouffer and colleagues describe Projective Large Deformation Diffeomorphic Metric Mapping (Projective LDDMM), a computational technique to integrate dense 3D tissue level MRI data with sparse measurements from histological or other optical imaging modalities. The authors demonstrate application through projection of neuropathological markers from histological images onto MRI data of the hippocampus.

Ionuţ-Gabriel Farcaş, Gabriele Merlo and colleagues developed a framework for uncertainty quantification and sensitivity analysis at scale by focusing on important input parameters. The framework was demonstrated to reduce computational effort and cost compared to standard methods in a turbulent transport simulation in the context of fusion research.

Roberto Torelli and colleagues propose a numerical framework to characterize fuel injection in internal combustion engines at multiple length and time scales. The approach demonstrates potential for increased fidelity in the flow dynamics by means of an affordable end-to-end methodology that links realistic injection operation to fuel combustion and engine emissions.

Shengnan Wang and colleagues report a digital twin framework of electrical tomography for quantitative imaging of a gas-liquid multiphase flow. This framework enables precise flow profile imaging using low-cost and noninvasive tomography techniques and can be extended to biomedical, aerospace and energy applications.

Bairaktaris and colleagues developed a printable flexible photodetector using a simple, scalable fabrication process. The photodetectors were demonstrated in practice in an augmented paper system. This low-cost technique could be also applied in robust and large-scale user interfaces.

An imaging device inspired by insect stereopsis allows for simultaneous near-distance microscopic imaging, high speed imaging at far distance and 3D depth imaging at intermediate distances. The camera, reported by Kisoo Kim and colleagues, gives clues as to how insects see the world and offers insights for designing compact cameras with multifunctional capabilities.

Shen and colleagues reported an unsupervised generative adversarial network (GAN) to identify patterns in leaves associated with superior mechanical properties and use 3D printing to build architected materials inspired by the patterns. In the future, this approach may be applied more broadly to natural materials to enable efficient algorithmic construction of structures with customized properties and form factors.

Fixed wing drone flight in dense (urban or forest) environments is challenging due to a need for a large area to turn. Inspired by the avian wing morphing, Enrico Ajanic and colleagues proposed and tested a drone with wings capable of folding and pitching, and a tail capable of folding and deflecting as a strategy to increase the roll moment, lift force, and reduce the turn radius. This finding enables agile drone flight within limited space.

Abdel-Rahman and colleagues introduce a discrete modular material-robot system that is capable of serial, recursive (making more robots), and hierarchical (making larger robots) assembly. This is accomplished by discretizing the construction into a feedstock of simple primitive building blocks combined with an algorithm to plan the optimal construction path and assemble the building blocks into functional units and swarms.

Thomas Matarazzo and colleagues determine modal frequencies of a suspension bridge and highway bridge from analysis of smartphone datasets during vehicle trips across the bridge. The field results suggest that data from both controlled and uncontrolled crowdsourced smartphone datasets have value in monitoring transportation infrastructure.

Berkey and colleagues quantitatively characterized partial-thickness cutaneous injuries after impact from projectiles simulating ballistic fragments. A corresponding damage model was developed to simulate and predict the cutaneous damage from impact, which could guide protective equipment design and clinical treatment.

Sozos and co-workers present and numerically evaluate photonic neuromorphic hardware using recurrent optical spectrum slicing for use in ultra-fast optical applications. The approach extends optical signal transmission reach to more than four-fold that of two state-of-the-art digital equalizers and reduces power consumption tenfold.

Optoacoustic tomography (OAT) is a hybrid imaging modality that combines optical excitation with ultrasound detection. However, OAT is limited by the lack of available high resolution ultrasound detection systems. Here Hahamovich and colleagues report single-detector OAT capable of high-fidelity imaging using an amplitude mask in planar geometry coded with cyclic patterns for structured spatial acoustic modulation.

Jan Kloppenborg Møller & Goran Goranović and colleagues introduce a data-driven twin methodology which balances physical knowledge with uncertainty quantifications. The approach makes it suited to application of real world problems with inherent unknowns. They demonstrate its application in the modelling and control of membrane water ultrafiltration

Kasidit Toprasertpong and colleagues describe reservoir computing hardware with potential for on-chip integration with existing computing technologies. The approach is based on a ferroelectric field-effect transistor, and can solve computational tasks on time series data including nonlinear time series prediction after training with simple regression.