MSVSCC 2019 THE 13TH ANNUAL MODELING, SIMULATION & VISUALIZATION STUDENT CAPSTONE CONFERENCE April 18, 2019 VMASC, Suffolk, VA ABM, Education General Sciences Infrastructure Security & Training & Engineering & Military Gaming Transportation, Medical & Virtual Reality Business & Industry Simulation PROCEEDINGS MSVSCC 2019 PROCEEDINGS Old Dominion University Department of Modeling, Simulation & Visualization Engineering (MSVE) and the Virginia Modeling, Analysis and Simulation Center (VMASC) held the 13th annual Modeling, Simulation & Visualization (MSV) Student Capstone Conference on April 18, 2019. The Conference featured student research and student projects that are central to MSV. Also participating in the conference were faculty members who volunteered their time to impart direct support to their students’ research, facilitated the various conference tracks, served as judges for each of the tracks, and provided overall assistance to the conference. Appreciating the purpose of the conference and working in a cohesive, collaborative effort, resulted in a successful symposium for everyone involved. These proceedings feature the works that were presented at the conference. Sincerely, Dr. Yuzhong Shen Capstone Conference Chair [email protected] Daniel Perez Capstone Conference Student Chair [email protected] MSVSCC 2019 KEYNOTE SPEAKER ANGEL SAN JOSE MARTIN received his Master of Science in Operation Research by the Naval Postgraduate School in Monterey, CA, in 2001. He is the Head of the Science and Technology (S&T), Modelling and Simulation (M&S) Section of the NATO Headquarters Supreme Allied Command Transformation in Norfolk, VA. He is a NATO civil servant and former Spanish Navy commander. He is responsible for the M&S activities on the Allied Command Transformation (ACT), as well as the coordination with the Science and Technology Organization of NATO. KEYNOTE: Simulating Alliance Security. The Movie. Modeling and Simulation opportunities in defense and security. Hot topics for the Alliance and allies that require attention from the M&S community. Beyond training, focus nowadays is expanding to M&S in support of analysis and wargaming to enhance decision making in NATO. Making M&S, AI, Data Sciences and optimization methods to work together is an essential element for innovation in NATO. Experts are required and opportunities will emerge. MSVSCC 2019 BEST PAPERS PER TRACK The tracks that the papers were divided up into included the following: • Agent Based Modeling, Education & Training • Gaming & Virtual Reality • General Sciences & Engineering • Infrastructure Security & Military Application • Medical Simulation • Transportation, Business & Industry For each track there were awards given out: The Best Paper award, and the Best Presentation Award. Those recipients for best paper are listed below by track. ABM, Education and Training Christine Odenwald, Germaine Badio, Jaron Stevenson, Travis Sullivan, Peter Marchione, Elise Feldt and Zhaohui Hu. “Virtual Autonomous System Training (VAST) Environment” Gaming & VR Shan Liu, Yuzhong Shen and Jim Leathrum. “A Review of Modeling and Simulation Tools for Autonomous Vehicles” General Sciences & Engineering Erik Jensen, Evan Coleman and Masha Sosonkina. “Randomized Asynchronous Linear Solvers” Daniel Perez. “Capsule Networks: Review of the Structures and Visualization of Features” Infrastructure Security & Military Application Yan Lu, Jonathan Graham and Jiang Li. “Deep Learning Based Malware Classification Using Deep Residual Network” Medical Simulation Haben Yhdego, Nahom Kidane, Dr. Rick McKenzie and Michel Audette. “Deep Learning Method for ECG-based Virtual Pathology Stethoscope Tracking in Patient Heart Auscultation” MSVSCC 2019 GENE NEWMAN AWARD Overall Best Paper – The Gene Newman Award The overall best paper is awarded the Gene Newman award. This award was established by Mike McGinnis in 2007; the award is given for overall best presentation, best paper, and research contribution. The Gene Newman Award for Excellence in M&S Research is an award that honors Mr. Eugene Newman for his pioneering effort in supporting and advancing modeling and simulation. Mr. Newman played a significant role in the creation of VMASC by realizing the need for credentialed experts in the M&S workforce, both military and industry. His foresight has affected both the economic development and the high level of expertise in the M&S community of Hampton Roads. The Students receiving this award will have proven themselves to be outstanding researchers and practitioners of modeling and simulation. For the 2019 Student Capstone Conference, The Gene Newman Award went to: Yan Lu, for her paper entitled ‘Deep Learning Based Malware Classification Using Deep Residual Network’ MSVSCC 2019 PROCEEDINGS General Sciences & Engineering PAGE 1 Gaming & Virtual Reality PAGE 50 Medical Simulation PAGE 65 Agent Based Modeling, Education & Training PAGE 81 Infrastructure Security/Military Application PAGE 113 Business, Industry & Transportation PAGE 132 MSVSCC 2019 PROCEEDINGS General Sciences & Engineering Applying a Test and Evaluation Framework to an Unmanned Maritime System Page 2 Ntiana Sakioti, Andrea Robey, Thomas Laverghetta, Nathan Gonda and James Leathrum A Game Theoretic Model of Pathogen Prevalence as an Outcome of Strategic Page 12 Interactions Between a Bacterium and a Tick Vector Sara Simmons and Holly Gaff Band Selection in Multispectral Images for Seagrass Detection Using Deep Convolutional Neural Network Page 14 Kazi Aminul Islam, Daniel Perez, Victoria Hill, Blake Schaeffer, Richard Zimmerman and Jiang Li Verbal Instruction to Perform Specific Running Styles Impacts Reaction Forces Page 16 During Habitual and Non-Habitual Running Collin Bowersock, Steven Morrison and Daniel Russell Embossed Construction and Explorative Visualization of 3D Structures from Page 28 2D Images Ahmet Saglam Anisotropic Mesh Adaptation Pipeline for the 3D Laminar Flow over a Delta Page 30 Wing Christos Tsolakis and Nikos Chrisochoides Capsule Networks: Review of the Structures and Visualization of Features Page 38 Daniel Perez APPLYING A TEST AND EVALUATION FRAMEWORK TO AN UNMANNED MARITIME SYSTEM Ntiana Sakioti Andrea H. Robey Thomas Laverghetta Nathan Gonda James F. Leathrum, Jr. Department of Modeling, Simulation, and Visualization Engineering Old Dominion University 5115 Hampton Boulevard Norfolk, VA, USA [email protected], [email protected], [email protected], [email protected], [email protected] ABSTRACT In order for an autonomous system to be confidently deployed in real world scenarios extensive behavioral testing must be conducted to ensure reliability. This is generally accomplished through the use of simulation followed by or in conjunction with testing in a physical environment. In this paper, the conceptual integration of an unmanned surface vehicle’s software to a test and evaluation framework is presented. The utilization of the suggested framework in this application enables testing throughout the virtuality-reality continuum, enabling behavioral testing early in the lifecycle, seamlessly integrating autonomous software with either a test environment or the physical vehicle. This reduces safety concerns as well as cost. This paper considers the utilization of the framework by mapping autonomous software developed for the Maritime RobotX competition to the framework and discussing the potential benefits of this activity. Keywords: autonomous systems, virtuality-reality continuum, test and evaluation framework, unmanned surface vehicle 1 INTRODUCTION Due to their nature and expected behavior, autonomous systems introduce various challenges in testing and validation. Expected behavior must be studied for occurrence reliability, while the system must safely adapt to external factors and possible complications. Throughout the development cycle, testing is performed in both virtual and physical environments, introducing inputs and models to improve autonomous software response to both expected and unexpected inputs. Utilization of a test and evaluation (T&E) framework is proposed so that system and subsystem testing can be conducted with augmented virtual and real data. This is accomplished by isolating autonomous software components from each other as well as from the operating environment, and ensuring that the autonomous software does not know the source of information. Isolating components and data enables testing to be conducted throughout the virtuality to reality spectrum, increasing the variety of tests and building confidence in the autonomous software (Leathrum et. al 2018a.). MSVSCC 2019, April 18, Suffolk, VA, USA; ©2019 Society for Modeling and Simulation (SCS) International 2 Sakioti, Robey, Laverghetta and Leathrum The autonomous software that will utilize and be mapped onto the framework provides the autonomy to an unmanned surface vehicle (USV). The USV’s software structure was designed around functionalities needed to complete the Maritime RobotX competition tasks (Maritime RobotX n.d.). Essential capabilities that the surface vehicle should demonstrate are listed below: • Obstacle avoidance • Navigation and control (circling an object, traversing through a gate, docking etc.) • Object and symbol detection (buoys, totems, docks, a hole cutout etc.) • Underwater beacon identification • Station keeping (maintaining current position) The remainder of the paper is organized into five sections. Section 2 details the framework that the authors propose for use in USV testing, while Section 3 presents