Dr. Kevin Gluck, US Air Force Research Laboratory, USA
Prof. Hong Hao, Curtin University, Australia
Prof. Sung-Bae Cho, Yonsei University, Korea
Dr. Komain Pibulyarojana, T-Net, Thailand
Prof. Seiya Ueno, Yokohama National University, Japan
Dr. Michael J. Leggieri, Jr., US Department of Defense, USA
Lieutenant General Norihiko Yamada, MD and Ph.D, Vice President of National Defense Medical College (NDMC), JSDF
Title: Human Performance Prediction: Basic Mechanisms and Exploratory Applications
Speaker: Dr. Kevin Gluck
It is standard practice worldwide to train based on one-time task checklists and fixed calendar-based refresher training intervals. These antiquated approaches make invalid assumptions about the dynamics of human learning and memory, ignoring most of what we have learned on those topics in the last century. These are risky, costly, and frustrating practices that should be discarded in favor of new methods informed by state of the art cognitive science and technology.
At the U.S. Air Force Research Laboratory we have been pushing the boundaries of cognitive science and technology to improve our ability to make evidence-based predictions of human performance. This resulted in a new patented technology called the Predictive Performance Optimizer (PPO). A key application of this technology is in the scheduling of learning events, for instance to acquire initial proficiency, to maximize retention, or to enhance relearning and maintain proficiency at a minimum required level. In this presentation I will describe key phenomena from the relevant empirical literature, criteria for the application of models of learning and memory, and initial results from a field study evaluating the use of PPO in training cardiopulmonary resuscitation (CPR), basic lifesaving skills relevant to military operations and more generally to medical and civilian communities worldwide.
Title: Discussion and improvement of concrete constitutive models for high-rate response predictions of RC structures
Speaker: Prof. Hong Hao (Curtin University, Australia)
With the development of computer hardware and computational mechanics techniques, numerical predictions of responses of reinforced concrete structures subjected to blast and impact loads have been becoming more and more popular for structural response analysis. The accuracy of the numerical simulations depends on the accurate representations of the structural material properties. A few concrete material models are available in literature, and are commonly used in simulation of structural responses. These models however differ from each other because they were developed by different researchers with respective available testing data. This paper compares these widely used concrete constitutive models. Discussions on their similarity, differences, and suitability in predicting the structural responses to blast and impact loadings are given. All these models assume concrete material is homogeneous and isotropic. To examine the influences of these assumptions, a series of uniaxial and triaxial tests have been performed. Experimental results indicate that concrete suffers obvious damage if the applied hydrostatic pressure is higher than the uniaxial compressive strength of concrete. This is a damage state overlooked by all the major concrete constitutive models. A mesoscale concrete model with the consideration of randomly distributed aggregates and pores is developed to explore concrete damage mechanism under hydrostatic pressure. The simulation results show that the stresses inside the concrete specimen are not evenly distributed under hydrostatic pressure, and existence of the deviatoric stresses is the primary cause of the concrete damage under high hydrostatic pressure. Discussions on the possible improvement on the current concrete material constitutive models by taking into consideration the concrete material behaviours under complex stress states are provided.
Title: Exploiting Machine Learning for Malicious Attack Detection in Cyber Security
Speaker: Prof. Sung-Bae Cho (Yonsei University, Korea)
Cybersecurity is defined by the technologies, processes and practices designed to protect networks, computers, programs and data from attack, damage or unauthorized access. One of the most problematic elements of cybersecurity is the quickly and constantly evolving nature of security risks in many organizations. They can share or release their data to third parties or even through the Internet; this can attract attention from malicious entities inside and outside the organization and pose serious threats on privacy and confidentiality.
Access control mechanisms are not enough to mitigate and defend from malicious attacks, and an intrusion detection system has been an active area of research in fields such as host and network systems, such setups have been found to be inadequate, ineffective, and unsuitable in detecting attacks. There is also the danger posed by insider threats. A legitimate user with malicious intent is a more serious threat than a user with limited privileges.
For this attack detection task, we exploit machine learning techniques to detect the malicious attacks. Machine learning methods are not only suitable in modeling normal behaviors, but also effectively minimize false-positive and false-negative rates. We present three methods such as random forest, learning classifier system, and evolutionary reinforcement learning that can accurately detect the attacks compared to other state-of-the-art techniques in machine learning, even on very skewed data. We will show the experimental results on the role intrusions against which the databases are operated with a role-based access control model.
Title: How SME company can be involved in the game changing technology of research and development in Defense Technology
Speaker: Dr. Komain Pibulyarojana (T-Net, Thailand)
In the present world, there are many terrorist risks, unexpect war, cyber attack which can identify or unidentified. The rapid change of technology, the speed of information disseminated on social network make us need to speed up of research and development in order to keep up with incident response and risk. One incident in one country can be happened as the same incident in every counties by the usage of social media. Because of the game changing technology, cyber security, national security and defense technology are now concerned as the same issue. The start-up company, spin-off company and SME will have chance to solve or response for threats.
The author had experience in information security more than 15 years. He worked as a leader of Information Security researcher in government agency. He founded ThaiCERT (Thai Computer Emergency Response team) and was the founded member of APCERT. He set up spin-off company(T-NET Co.,Ltd.) from research unit to have services in IT security and Defense Technology. He will share how SME can have fast services and set up the linkage between the Cyber Threat, National Security Threat and Fast research for in time product.
The research and development’s roadmap was started with mobile telephone Jammer, remote control’s frequency Jammer step to Drone Jammer which prepare for the drones are in widespread use, more and more both in beneficial and malicious purposes. Violation of liberty and rights, espionage, spying, terrorism, etc are a few examples of malicious uses. The example research of Drone Jammer will help protect community from abusing drones and show how SME have chance to be involved in the game changing technology.
Title: Guidance Law for Multiple Aircraft Based on Optimal Control Theory
Speaker: Prof. Seiya Ueno (Yokohama National University, Japan)
Necessity of guidance law for multiple aircraft has been increasing in the recent years. For example, guidance law of multiple missiles attacking one target simultaneously increases the success rate of mission, or air traffic management of multiple aircraft approaching one runway decreases the delay of schedule with high safety. Optimal control theory can be applied to such type of problems and many papers have been written by many researchers. The results of papers, however, have not applied to the actual systems, because the computational load for multiple aircraft is very high and the stability on convergence speed of iterative calculation is poor due to the nonlinearity of problems. Such disadvantages are hard wall for application of optimal control theory to the real guidance systems.
The author has focused on the application of optimal control theory for many years. The results of recent researches on development of real-time optimization techniques are presented in this paper. The first example is utilization of a universal chart, which is calculated prior the mission and provides the optimal trajectories for arbitrary initial condition without iterative calculation. The examples of simultaneous attack by multiple missiles are shown. The next example treats the minimum time problem for reconfiguration of formation flight by multiple aircraft. The calculation technique in this example is based on the feature that the necessary conditions of minimum tine problem with specified position is the same as the necessary conditions of maximum or minimum range problem with specified terminal time. Thus the reachable area of aircraft can be calculated prior the flight, and the terminal time is searched during flight. The searching procedure can be executed by each aircraft, thus the total calculation time is short. The examples of reconfiguration of multiple aircraft are shown.
Title: Scope of Blast Injuries
Speaker: Dr. Michael J. Leggieri, Jr. (US DoD, USA)
Explosive weapons remain the weapons of choice by our adversaries. The injuries these weapons cause span a broad spectrum of blast injury types based on injury mechanisms from primary through quinary, and severities from mild to severe. Blast injuries are typically complex, with more than one type of injury occurring at the same time. Injury data from recent conflicts confirms the diversity, complexity, and severity of blast injuries. Medical research programs which focus on blast injury prevention and treatment are equally diverse, cutting across research organizations and scientific disciplines. In 2006, the US Congress addressed the need for a more coordinated approach to addressing complex blast injury issues by directing the establishment of a DoD Executive Agent (EA) who would be responsible for developing a coordinated and responsive DoD blast injury research program capable of delivering timely and effective blast injury prevention, mitigation, and treatment strategies. This presentation will describe the EA’s blast injury research coordination efforts within the context of the types and complexity of blast injuries observed in recent conflicts. It will highlight the challenges in coordinating efforts across research organizations and diverse medical and non-medical communities. It will also describe the EA’s key research coordination initiatives which are focused on advancing our understanding of blast injuries, and accelerating the development and fielding of effective blast injury prevention and treatment strategies to better protect Service Members from blast injuries, and care for those who are injured.
Title: The possible alteration in research landscape of defense medicine
Speaker: Lieutenant General Norihiko Yamada, MD and Ph.D (NDMC, JASDF)
National Defense Medical College (NDMC) was established in 1973 to alleviate the serious shortage of medical officers of Japan Self Defense Force. At first, the focus of education and research at NDMC was almost faithfully modelled after that of civilian counterparts, so that it did not necessarily put major stress upon medical areas relevant to military affairs. It was not until the post-cold war era and especially 1995, when we almost coincidentally experienced a devastating earthquake at Kobe-city of 1 million people and Tokyo subway sarin gas attack incident, that we realized the significance of military and disaster medicine (Defense medicine in a word). NDMC Research Institute for defense medicine was established in 1996. Since then, its organization and research agendas have been arranged in line with ever-changing research requirements which mainly reflected key medical issues including so called ‘signature wounds’ such as PTSD*, or m-TBI*. Ministry of Defense recently adopted the research program called ‘Advanced Research on Defense Medicine’. Thanks to the program, we could farther enhance representative research activities, such as blast injury. At current presentation, I would briefly introduce our institute and some examples of its recent achievements.
We have been keen in reacting to and studying about ‘signature wounds’ of recent but past incidents. It has been a successful business in that the approach has spawned various kinds of medical spin-offs. Notably, the people at the time could often recognize the ‘signature wounds’ even during the incident. This means that the ‘wounds’ was not of unknown category, but was already known or established at the time. And history tells us that unawareness of the ‘wound’ among 1st responders, medical personnel and other professionals repeatedly caused the inadvertent extension of human damages. It could be said that we have not incorporated existing relevant medical knowledge and skills into military and/or disaster preparedness and/or response well enough. If the goal of our research is to reduce the human damage in future, I think ‘prediction of the next signature wound’ should be treated as a more serious and feasible research agenda than ever. To confront the issue, we might need a totally different approach from traditional medical research style and organization. At this presentation, I would also refer to a few examples which might imply the new research landscape even in traditional research areas. It now seems that considerable alteration in research landscape of defense medicine is not only indispensable but is already under way.