Year: 2015

Authors: Shane Gallagher, Brenda Bannan, Shelly Blake-Plock, Bridgette Lewis

Abstract: As a result of next-generation networking and the Internet of Things (IoT) technologies, big data analysis is possible and has been shown to have a positive impact on areas of national significance yet requires new tools to deal with the variety and quantity of data multiplying at an exponential rate.. Concurrently, IoT technologies are rapidly becoming a mainstream data source. Training simulations have historically been limited either to computer-based simulations or live human-observable field-based simulations ;however, IoT technologies can open up innovative, hybrid digital-physical opportunities both for delivering and for understanding the outcomes of training in a much more dynamic and comprehensive way. The feasibility of IoT technologies in training has historically been limited by interoperability and scale. However, Advanced Distributed Learning’s Experience Application Programming Interface (xAPI) allows interoperability and scale in next-generation training environments and provides a way to standardize the formative data of human experience captured through digital context. It also provides a way to capture information and formalize human experience from multiple and varied networked devices into standardized, human-readable statements. These can inform both human and machine learning through leveraging big data analysis and interoperability of the IoT technologies. By leveraging the xAPI and IoT technologies as a cyberphysical system embedded in virtual and live training scenarios, it is possible to capture and measure real-time team performance for immediate analysis and remediation or for post hoc analysis in after action reviews. This paper discusses the application of learning analytics and design for an IoT context through describing the implementation of 1) a live action medical simulation as part of the Global Smart Cities Challenge (sponsored by the NIST and the OSTP) and 2) the proposed capture and analysis of communication performance data and measures within specific coalition training scenarios supporting the 2015 Bold Quest Assessment sponsored by the Joint Fires Division of the Joint Staff.

Year: 2015

Authors: Brian Sanders, Brent Terwilliger, Ken Witcher, Mark Leary, James Ohlman, Christina Tucker

Abstract: Offering laboratories and team projects present significant challenges for delivering Science, Technology, Engineering, and Mathematics (STEM) courses in the online (asynchronous) modality. These interactive workspaces are important attributes since they provide forums for students to more deeply explore fundamental principles, exercise teamwork and planning to jointly overcome problems, and gain critical experience. The employment of online environments and interactive activities hold the potential to change how fundamental student outcomes measured by accreditation organizations are incorporated and treated in curricula, potentially improving the quality of the overall educational experience. To address this need Embry-Riddle Aeronautical University has teamed with Pinnacle Solutions to develop a realistic unmanned aircraft system (UAS) development, application, and evaluation simulation that educators can integrate into program curriculum. The research contained in this paper addresses simulation development and application starting with identification of basic educational objectives driving the need and how the simulation tool is envisioned to satisfy learning objectives. This will be followed by a description and examples of a multi-environment simulation framework designed to meet those needs. The first is a component test environment where students can investigate basic technical principles of operation and key performance metrics of standalone UAS components such as sensors, communications, and propulsion elements. The second is an integration facility, where students are provided the capability to apply knowledge gained in the previous laboratory to select and combine appropriate elements into a unified subsystem to meet prescribed mission parameters. The third is a flight test environment, where students experiment with development and execution of simulated flight profiles over common terrain environments (i.e., mountainous) to measure operational performance attributes of the completed UAS. The design is anticipated to provide the flexibility to implement each environment sequentially, as described above, or independently; ensuring a solution applicable to a broad range of courses, objectives, outcomes, and student capabilities.

Year: 2015

Authors: Brian Stensrud, Charles Newton, Beth Atkinson, John Killilea

Abstract: Over the past decade, we have seen moderate demand for simulation-based training systems to include automatic speech recognition (ASR). Like commercially available services such as Apple's Siri and Google Now, ASR gives training systems the capability to interpret human speech and react to that speech with appropriate actions (e.g. executing a spoken command) and responses (e.g. replying to a human with confirmation or requests for clarification). Introduction of this capability is designed to address instructor-manning limitations and improve the fidelity of the training experience. However, ASR successes within simulation-based training systems have been modest, historically. We contend that this lack of widespread usage and success stems primarily from a fundamental misunderstanding of (and thus lack of investment in) the components necessary to achieve more effective ASR. In this paper, we describe the essential functions of ASR: (1) Recognition is when the audio of the spoken utterance is translated into text. (2) Understanding attempts to glean meaning from the text – whether they denote, for example, a new directive, a response to a previous query, or a request for new information. (3) Behavior refers to the functions the system is responsible for after receiving a recognized speech utterance. (4) Some training systems also employ dialogue when continuous interaction with humans is required. Finally, we outline current ASR research and development, discuss typical implementations, and introduce potential strategies to improve specific ASR functions and the capability as a whole to provide better support for future training systems.

Year: 2015

Authors: Lloyd Werk, Maria Carmen Diaz, James Franciosi, Tim Wysocki, Lorie Ingraham, James Crutchfield

Abstract: Military and civilian healthcare is undergoing radical transformations in almost every aspect of patient care from diagnosis to treatment. Along with increased complexity in the technology of delivery systems and procedures, medical knowledge is expanding at an ever-increasing rate, and yet clinicians are expected to retain knowledge and remain proficient in their fields. Frequency of exposure to specific clinical problems and processes are known contributors to physicians’ decay of clinical knowledge and proficiencies. For example, while deployed, military physicians may experience less demand for specific clinical skills and are, therefore, at risk for knowledge decay. A systematically applied knowledge retention program integrated with continuous training is one possible response. However, institutionalizing standardized training at fixed intervals for all may not be the most cost-effective nor efficient solution. This paper discusses the progress of a research study tasked to develop and validate efficient interventions to mitigate physician knowledge decay that address both increased domain complexity and lower frequencies of exposure. The process of intervention selection is based on the analysis of elements of the care for nine targeted clinical problems that reveal physician knowledge decay with decreasing frequency of exposure to those clinical problems. Once the most critical elements of the care process have been identified, we apply a structured approach for selecting, developing, and evaluating possible interventions geared towards choosing those that specifically address identified knowledge needs and align with the organization’s learning goals, infrastructure and operating budgets. Recommendations for a systematic, yet flexible, method for evaluating, weighing and scoring multiple knowledge decay mitigation alternatives are included, supporting interventions ranging from static job aids to immersive learning simulations. In summary, this paper proposes a comprehensive selection model for continuing medical education programs committed to prevent skill decay, aid knowledge retention and improve overall physician and organizational performance.

Year: 2015

Authors: Steven de Jong, Wouter Noordkamp, Nick van der Poel, Selmar Smit

Abstract: Assessing the effectiveness of a plan, given multiple potential scenarios, is a common problem for analysts, especially in the military domain. This problem can seriously impact the safety of the people that are involved in planned missions. More precisely, the availability of multiple models, with varying levels of fidelity, leads to the complex task of selecting the best model(s) to assess the effectiveness of a plan. Under time constraints, optimal model selection depends not only on the fidelity of the models at hand, but also on the nature of the possible scenarios the plan applies to, such as the potential presence of stochastic variables and the number of different scenarios that have to be evaluated in order to obtain a reliable estimate of the true effectiveness of the plan. In this paper, two algorithms are presented to maximize the reliability of the obtained plan effectiveness under time constraints. To this end, the algorithms select the best model(s) as well as the most appropriate scenarios. Both algorithms have been tested on syn-thetic data as well as on two Navy-related use cases. Results show that both algorithms reach a higher level of reliability within the given amount of time than conventional approaches. Thus, they allow analysts to better assess the effectiveness of their plans and therefore they increase the safety of everyone involved in planned missions.

Year: 2015

Authors: Lisa Tripp, Elliot Humphrey, Christine Covas-Smith, Jonathan Diemunsch, Mike Garrity, Cullen Jackson, Mike Keeney, Sterling Wiggins

Abstract: Aircrews have leveraged simulation for several decades to immerse themselves in complex mission situations, develop new concepts of operations (CONOPS) and tactics, techniques, and procedures (TTPs) for the next fight, test new capabilities, and develop robust and adaptable mission performance. While the Air Force endorses simulation-based training as a vital need for aircrews, currently there is not an analogous capability available for intelligence analysts. This gap becomes more crucial as we prepare the next generation of analysts for potentially drastically different operational environments where Air Supremacy is not guaranteed, denied environments are the norm, and cyber warfare plays a frightening role. Simulation-based training is needed to prepare analysts for these environments. Although, in general, simulation-based training is thriving across many domains (e.g., flight simulation, driving simulation, shooting simulation), little work has focused on training for Intelligence, Surveillance, and Reconnaissance (ISR) professionals. Requirements for developing a realistic, simulation-based training environment for ISR tasks are distinctive from those required for Aircrews. Simulation of the large variety of information sources is the key. The objective of the current effort was to identify the requirements underlying development of a simulation-based training capability to maximize the efficiency and effectiveness of training for ISR. To tackle this complex problem set, the Air Force Research Laboratory (AFRL) Warfighter Readiness Research Division leveraged Mission Essential Competency analysis in conjunction with cognitive task analysis to identify key requirements for high fidelity, simulation-based training for the ISR. This paper will describe the challenges facing creation of simulation-based training for ISR, work to develop a training capability for these critical warfighters, and our vision for future ISR simulation-based training.

Year: 2015

Authors: Logan Williams, Charles Bullock, Marc Winterbottom, James Gaska, Steven Hadley, Charles Lloyd, Michael Browne

Abstract: Game-based flight simulation software has been shown to provide a reliable, low-cost, virtual environment able to facilitate a wide range of training and research objectives. In this work, which is part of the U.S. Air Force School of Aerospace Medicine Operational Based Vision Assessment program, game-based simulation software was used to render an immersive three-dimensional constructive environment within a helmet mounted display (HMD) for weapons platform specific vision research and to quantify the impact of aircrew vision on selected operational tasks. In this work, an operationally relevant MH-60 call-to-landing task was simulated to provide data relevant to the applicability of U.S. Air Force Flying Class III depth perception standards. The specific simulation system consisted of a high-resolution (1920x1200) 55° field-of-view binocular HMD with infrared head tracking, in which two instances of X-Plane were stereoscopically rendered to the HMD using separate PCs, both incorporating Intel i7 processors and Quadro K4200 video cards with Quadro Sync. This paper details the overall design, implementation, and validation of the virtual environment used to simulate the MH-60 call-to-landing task, including stereoscopic rendering using game-based simulation software, hardware/software stereo rendering limitations, HMD warping, and head-tracker integration. The minimum perceived stereo threshold capabilities of this system are also quantified, including discussion of its applicability to simulated tasks requiring precise depth discrimination. This work will provide an example simulation framework for future stereoscopic virtual immersive environments applicable to both research and training.

Year: 2015

Author: R. Scott Starsman

Abstract: Traditional approaches to 3D scene reconstruction require very large image sets, are extremely processor intensive, and perform poorly when faced with surfaces with limited features. The work described in this paper builds upon an approach presented at I/ITSEC 2013 that greatly increases quality of 3D reconstruction, requires a minimal set of images, reduces model storage size, and is resilient in the face of low-feature surfaces. While the work presented in 2013 demonstrated the successful reconstruction of a 3D model from a small set of images, it suffered from several problems including long processing time, low success rate for arbitrary structures (meaning that it worked well on specific types of buildings/structures but not on the vast majority), sensitivity to misidentification of commonly present elements, such as logos and signs, and correspondence point noise. This paper details the methods used to address these deficiencies and achieve the full promise of rapid scene reconstruction in the face of a limited number of images. Two elements of the image processing pipeline were identified that led to the performance issues and replacement algorithms and processes were developed and integrated into the model. The replacement of those components dramatically improved performance and supported the generation of arbitrary structures from an image set. Key issues that have been resolved include: extremely long processing times, sensitivity to structures with few surface features, sensitivity to repeated features, and sensitivity to correspondence point noise. A description and derivation of these new approaches is discussed and as a final demonstration, the system is used to generate a 3D reconstruction of a city block with the results capable of being viewed in a tool such as Google Earth. This work is pertinent in the military, security, simulation, and disaster response scenarios.

Year: 2015

Authors: Robert Pokorny, Jacqueline Haynes, Lisa Holt, Michael diPilla

Abstract: Navy maintenance is becoming increasingly difficult with more complicated systems, reduced staffing, and efforts to reduce expensive training time. To improve maintenance readiness, the Navy is testing tablets by which technicians receive performance aids and directly connect with the larger Navy maintenance enterprise system. Technicians’ performance can be improved by (1) micro controlled experiments which investigate interface details of accessing and presenting content via tablets and (2) macro field tests that illustrate the effect of technological tools in their deployed state. A comprehensive approach to improve productivity and decrease cost through tools is best informed by both micro and macro studies, and to integrate the results of both to create and promote Navy goals. The micro element of our approach is the study of how a tablet-based presentation of procedures can be structured to provide technicians the support needed to maximize performance. We will report results from one such study, identifying difficulties introduced by tablets and how they can be overcome, and the capabilities now possible with interactive tablets. The macro element of our comprehensive approach is the study of how the Navy maintenance technicians can benefit when connected to enterprise resources. Technician benefits include an ability to order components when technicians are in the field, access updated technical documentation, and automatically collect work performance data which reduces redundant paperwork and enables big-data analytics to identify interesting trends of previously unknown efficiencies and performance difficulties. We will report results from a recent field test that includes lessons learned from connecting technicians to the enterprise system. Micro studies provide scientific verification of principles used to develop the solution, and macro studies reveal how well the solution improves work flow and productivity in the Navy context.

Year: 2015

Authors: Rosalyn Scott, Brian Burke, Nancy Benton, Helga Carabello, Mary Davidson, Paul Ingmundson, Sean McCoy, Cathy Graham, Terry Oroszi, Manny Dominguez, Jennie Gallimore

Abstract: The Veterans Health Administration is the largest integrated health system in the world serving Veterans in both urban and rural environments. To enhance clinical outcomes and education, a VA Virtual Medical Center (VMC) has been launched as a collaborative care and learning environment. Resources can be accessed anytime and anywhere. Capabilities include virtual clinics with the integration of current telehealth technologies, cybraries for patients and healthcare team members with electronic resources and searchable medical content, serious medical games, e-learning platforms and conference venues. A full range of learning technologies, including virtual patient and standardized patient-based platforms are fully integrated into the environment. Our implementation strategy leverages ways in which the VMC can be synergistic with existing care models; decrease repetitive staff activities; increase dissemination of and participation in educational interventions; provide more effective education; capture productivity measures; and, be easy to navigate. Input from human factors engineers, clinicians, educators, and technology experts has been critical. Initial implementation includes five pilot projects characterized by the need for educational interventions for patients or/and healthcare team members as well as clinical interventions to optimize Veteran health outcomes in key clinical areas. The clinical areas include diabetes, sleep disturbances, congestive heart failure management, obesity, and palliative care. Interventions include staff training for new protocols, peer and professional coaching for patients with chronic diseases, shared medical appointments, training and resources for uisng at home equipment such as CPAP machines. Assessment strategies are comprised of a global assessment of the technologies in place and project specific ones tracking outcomes. Our newest generation of Veterans is very tech savvy and embraces virtual world technologies. The VMC will allow geographically separated staff and patients to interact in a rich avatar-based environment. In-world opportunities can provide important care resources and rich educational experiences for all learners.