20 Aug 2020

PUBLICATION NOTICE: Three Rivers, Southeast Arkansas Navigation Study: Ship Simulation Report

Abstract: The McClellan-Kerr Arkansas River System (MKARNS) is a major inland waterway that begins at the Port of Catoosa in Tulsa, OK, and travels to the confluence of the White and Mississippi Rivers. Over the years, many structures have been built to help control overland flow between the White, Arkansas, and Mississippi Rivers. These structures have required a significant amount of rehabilitation, which has resulted in high maintenance costs. The US Army Corps of Engineers and the Arkansas Waterways Commission conducted the Three Rivers Southeast Arkansas Feasibility Study (also known as the Three Rivers Study). The Three Rivers Study focused on providing long-term dependable navigation in the MKARNS. From this study, a proposal was developed that included a 1,000 ft reopening of the Historic Cutoff and a reinforcement of several areas near the White River. In 2019, the US Army Engineer Research and Development Center Ship/Tow Simulator was used to perform a navigation study to ensure the proposed modifications did not negatively impact navigation on the White River section of the MKARNS. Assessment of the proposed modifications was accomplished through analysis of ship simulations completed by experienced pilots, discussions, track plots, run sheets, and final pilot surveys.

20 Aug 2020

PUBLICATION NOTICE: Understanding State-of-the-Art Material Classification through Deep Visualization

Abstract: Neural networks (NNs) excel at solving several complex, non-linear problems in the area of supervised learning. A prominent application of these networks is image classification. Numerous improvements over the last few decades have improved the capability of these image classifiers. However, neural networks are still a black-box for solving image classification and other sophisticated tasks. A number of experiments conducted look into exactly how neural networks solve these complex problems. This paper dismantles the neural network solution, incorporating convolution layers, of a specific material classifier. Several techniques are utilized to investigate the solution to this problem. These techniques look at specifically which pixels contribute to the decision made by the NN as well as a look at each neuron’s contribution to the decision. The purpose of this investigation is to understand the decision-making process of the NN and to use this knowledge to suggest improvements to the material classification algorithm.

11 Aug 2020

PUBLICATION NOTICE: The Urban Ground-to-Ground Radio-Frequency Channel: Measurement and Modeling in the Ultrahigh Frequency Band

ABSTRACT:  Ground-to-ground radio communication and sensing within the urban environment is challenging because line of sight between transmitter and receiver is rarely available. Therefore, radio links are often critically reliant on reflection and scattering from built structures. Little is known about the scattering strength of different buildings or whether such differences are important to the urban ground-to-ground channel. We tested the hypotheses that (1) diffuse scattering from built structures significantly impacts the urban channel and (2) scattering strength of urban structures varies with surface roughness and materials.  We tested these hypotheses by measuring urban channels in Concord, New Hampshire, and Boston, Massachusetts, and via channel-modeling efforts with three-dimensional representations of the urban environment. Direct comparison between measured and modeled channels suggest that both of these hypotheses are true. Further, it appears that ray-tracing approaches underestimate the complexity of urban channels because these approaches lack the physical processes to correctly assess the power incident on and scattered from built structures. We developed a radio-geospatial model that better accounts for incident power on both directly visible and occluded buildings and show that our model predictions com-pare more favorably with measured channels than those channels predicted via typical ray-tracing approaches.

11 Aug 2020

PUBLICATION NOTICE: Autonomous QUerying And PATHogen Threat Agent Sensor System (AQUA PATH): Monitoring Source Waters with Geospatially Wirelessly Networked Distributed Sensing Systems

Abstract: Contaminants serve as health risks to recreational water, potable water, and marine life that result in undocumented effects on population exposure. In many areas of the world, the concern lies in contaminated drinking water, which would immediately effect social and economic order. As research advances for innovative solutions, the deployment of automated systems for source water monitoring could reduce the risk of exposure. Water quality monitoring typically involves sample collection and analyses that are performed in a laboratory setting. These results are normally presented after an 18−48 hr period. This report details the prototyped Autonomous QUerying And PATHogen threat agent sensor (AQUA PATH) geoenabled system that is able to detect the presence/absence of pathogenic bacteria indicators in source waters and report these values in the field, in less than 30 minutes. The AQUA PATH system establishes rapid field data collection and reports assessment of source waters bacterial loads at near shore inner coastal locations, which makes a leap forward compared to current presence/absence tests standards established by the EPA.