ABSTRACT. To prevent and stop illegal and malicious maritime activity, it is necessary to better understand the people, places, organizations, and vessels that contribute to those activities. This work examines whether spatial-temporal co-occurrences are useful for finding connections among vessels and providing understanding of their activities. To identify those occasions, we used two datasets, Automatic Identification System (AIS) records of vessel positions, and time and company information on the ownership of vessels to confirm if vessels were affiliated. Three experiments studied co-occurrence among groups of vessels that included randomly selected vessels, vessels of interest, and locations of interest. Vessels were identified for each of the three experiments and their co-occurrences were identified. We found that co-occurrence manifested in different ways, and while it was not especially useful in demonstrating common ownership among vessels, it was helpful for other purposes.

ABSTRACT. Systems Engineering for complex and sociotechnical systems is becoming increasingly difficult due to growing complexity in the problem and solution space. This resulted to the development of Model Based Systems Engineering (MBSE), which applies modelling supported by an architecture framework, modelling language in a software tool. Modelling is applied to capture and represent the mental models of the systems’ stakeholders during the requirements analysis and concept development stages. The software tool enables capturing model information in a database through various diagrams and views. This data centric approach ensures consistency and traceability of the defined model elements and their relationships. Since Command and Control (C2) systems can be viewed as complex sociotechnical systems, MBSE should be useful for capturing requirements to support development of concept solutions. The MBSE method presented in his paper assists capturing of requirements for a C2 system as well as identification of the main logical block elements and their required functionality. This approach is demonstrated through development a concept C2 system solution suited to the complex operational environment of Special Operations.

ABSTRACT. A prerequisite for exercising Agile C2, and thereby achieving decision superiority over the adversary, is the effective management of information across tactical networks. This in itself is a highly complex and challenging problem due to the limitations imposed by the Disrupted Intermittent and Limited (DIL) tactical networks and the increasing demands for information placed on those networks.

In conjunction with the US Army Research Laboratory, Defence Science and Technology (DST) in Australia is developing an intelligent autonomous software agent, known as SMARTNet, that will attempt to address some of these difficult challenges. The idea behind SMARTNet is to introduce dynamic transformation, prioritisation and dissemination of tactical information based on a current military operational context and a current network state. The first iteration of the middleware, which was developed solely by DST, addresses the important use-case of dynamic prioritisation of Position Location Information (PLI) of friendly forces, in contention with other friendly tactical traffic. This paper outlines current land tactical needs and challenges in the information space and motivates the use of intelligent autonomous information management as a key part of any potential solution. We explore the methodology applied in developing our initial concept demonstrator and the experimentation performed to evaluate the PLI use-case. We present initial findings from the experiments, discuss the implications of these and conclude with recommendations for future work in the SMARTNet program.

ABSTRACT. During large command post exercises in the Canadian Armed Forces, training audience and staff are responsible for submitting lessons learned observations to a central collection. These observations form an important data source for the post-exercise analysis. One of the analysis goals is to find trends or common themes that emerge from these data. However, reading all the observations and defining common topics can be a very time consuming task for an analyst. Moreover, the required effort does not scale well as the size of the collection increases. To assist in the production of timely and scalable analyses, this paper demonstrates the use of topic modeling – a statistical method for discovering topics in a collection of documents. We discuss the chosen method and data cleaning issues. Then we compare topics generated by this method with those found in an existing analysis of observations from a large exercise at the Canadian Forces Warfare Centre. Faster discovery will help analysts study larger collections and identify topics as they emerge, allowing them to further investigate issues while the exercise is still running.

ABSTRACT. The technologic capabilities of autonomous systems (AS) continue to accelerate, and integrated performance by AS and people working together can be superior to that of either AS or people working alone. We refer to this increasingly important phenomenon as Teams of Autonomous Systems and People (TASP), and through our recent research—representing the current state of the art—we have demonstrated computational experimentation capability in the TASP domain. This stream of research seeks to stay five to ten years ahead of practice, which enables us to anticipate both issues and opportunities in an area that remains under researched: C2 of autonomous systems. In this technical paper, we build upon our computational experiments to address how TASP C2 can be designed for asymmetric advantage, with particular emphasis on dynamic design, through which an organization can increase agility and outperform otherwise peer or equivalent organizations. In particular, we conceptualize five C2 maturity levels and corresponding design archetypes, and we outline key conditions and transitions between them for military organization agility, using examples from combined manned-unmanned aircraft operations at sea and multinational disaster relief.

ABSTRACT. The aim of this study is to provide further insight into the concept of self-synchronization and to explore the possible implications of adopting this alternative approach to traditional hierarchical C2. The central question is: How to continue the function of command and control without hierarchical commanding of entities? First, the C2 function is explored, including its sub-functions (leadership, control, decision making and attune ) and underlying principles (continuity, clarity, unity and integration). Subsequently, self-synchronization is explored, as this could prove to be a useful concept to diminish hierarchy and to allocate more decision rights down to field level. According to Albert & Hayes (2003) self-synchronization is described as the operating of entities in the absence of traditional entities for C2. To assess self-synchronization and the implications for the function of C2, this study analyses self-synchronization in terms of characteristics, composition, enablers and challenges. Improved insights on how to adopt self-synchronization helps identifying new opportunities for future implementation of the C2 function, which will increase C2 agility. This research is conducted within the Dutch research program NetForce Command and the Multinational Capability Development Campaign (MCDC) Information Age C2 project, cycle 2017-2018. Both look at self-synchronisation as possible alternative approaches to traditional hierarchical C2.

ABSTRACT. Future military commanders will operate in a globally connected hybrid battle space, in which adhoc coalitions of state and non-state actors will perform tasks based on a combination of centralized and decentralized command structures. This operational environment will put high demands on the military organization from a technical, organizational and information provision perspective. In this paper, we aim to define the future set of requirements for IT systems that will need to support the C2 applications of the future. Based on this overview, specific functionalities that support the future military operational environment will be described. This set of functionalities serves as the basis for a more detailed specification of the future C2 support architecture. We start by describing the future user environment and the implications for the high level design requirements. Based on this description, the requirements of the C2 supporting architecture and functional design will be elaborated, after which a first concept of the networked command environment will be shown.

ABSTRACT. The proliferation of augmented reality (AR) and virtual reality (VR) systems has accentuated their potential utilization as strategic, tactical, and training platforms. However, little is known about the individual and team behavior associated with using AR in military contexts. In this study, two-member teams conduct a simulated mission planning scenario with AR technology (Microsoft HoloLens) and a multi-touch 2D display technology (Microsoft Surface table). Team members interactively manipulate a small subset of military symbology in the process of planning their mission to retrieve a repository of intelligence documents within enemy-held territory. Differences in individual behavior and team coordination are observed and compared across technologies. The results of this research could be used to inform models of AR behavior applied to specific steps in the military decision-making process.

ABSTRACT. To adapt the changes from the battlefield environment and the self-status, command and control (C2) system should have the ability of agility. Modelling the agile C2 system is still an open problem in the literature of military command and control. In this paper, the conceptual model of the agile C2 system was given by describing dynamics of change,capability characters, and operation evolution process. Then, based on the self-adaptive theory, the agile operating mechanism to cope with these changes was proposed.

ABSTRACT. The accelerating evolution of Fog Computing technologies led to a growing interest towards their applicability to the Internet of Battlefield Things (IoBT). Efficient design for Fog Computing applications need to consider the optimal use of the available resources at the edge and the Cloud, switching from one other depending on the current status, execution price, and user requirements. This is a complex task, as the optimization needs to consider the distributed and dynamic nature of the environment. There is the need to support investigation efforts by enabling researchers to experiment with Fog environments in a controlled and reproducible fashion. Unfortunately, most simulators do not implement a service model suited for Fog computing applications. This paper presents Phileas, a simulator that supports the definition of Fog services. Phileas allows to reenact the behavior of Fog services and evaluate different service policies and allocation strategies.

ABSTRACT. Although the use of the Internet of Things (IoT) in a smart city context primarily stems from civilian applications like management and monitoring, there are also other uses. For instance, military forces may utilize sensors during natural disasters, terrorist attacks or insurgent attacks, in order to improve situational awareness. For this to be possible, the military forces must be able to discover the sensors, and be able to utilize the data formats they use in their C2 systems. It is therefore necessary to be able to describe the sensors in a standardized and machine-readable way, such that they can be automatically discovered and their capabilities assessed.

In this paper, we investigate using open standards such as OpenAPI and SensorML to enable sensor sharing so that IoT can be leveraged by military forces in an agile and interoperable way, which is necessary to support coalition operations.

ABSTRACT. Future Command and Control (C2) need to be both agile and robust to cope with unpredictable operational environments and to make use of the technological advances. The Swedish Armed Forces develops a concept for future C2 of military operations. The objective of this paper is to explore the utilization of mission command in a future C2 concept. Sweden has a solid tradition of mission command, promoted by the Swedish culture of participation and empowerment. However, the future operational environment demands a need for organizational agility, thus changing the prerequisites for mission command. The future C2, which is under development, concept encompasses centralized and decentralized command, where forces have varying degree of self-sustainability. On one hand, technologically advanced exclusive resources need to be carefully prioritized. On the other hand, dynamic situations demand rapid decision making and seizing the opportunity given in the moment. The future operational environment includes hybrid warfare and gray zone issues, demanding thorough analysis in order to foresee the political consequences of decisions. Sensor and communication technologies enable enhanced situation awareness; however, the technical infrastructure is vulnerable. This paper discusses the implementation of mission command in a future C2 concept from a two-fold perspective: how will mission command contribute the conducting of future military operations and how should the socio-technical C2 systems be designed to enable agile decision-making? The conclusion is that mission command is still relevant in the complex future operational environment, however, the increasingly complex operational environment demands continuous development of the C2 function.

ABSTRACT. • Digital information sharing in ad hoc civ-mil teams is crucial for success in current and future missions. • This is hampered by current security policies, interoperability issues and security mechanisms and will not be solved by FMN in the near future. • TNO is working on a concept for civ-mil info sharing in dynamic adhoc teams, that addresses this issue in a quick, user-friendly and technically feasible way.

ABSTRACT. The world we live in is a complex system tinged with constant change. In order to cope with this fact, a defense mission system needs to adapt to these challenges. The command and control (C2) system is the component of the defense mission system that is our system of interest, our unit of analysis, in this paper. We present a first attempt to use a method based on design logic complemented with scenario driven exercises, to extract requirements and more fine-grained design criteria to enhance design of future C2 systems. As a starting point, three scenarios were developed that intended to reflect key features of future potential conflicts. A number of subject matter experts (SMEs) participated representing strategic, operational and tactical levels of command. The SMEs were asked to state their C2 requirements in each of the scenarios. The resulting set of C2 requirements were analyzed to find design criteria pertaining to the general C2 functions (Data Providing, Orientation, Planning, Influence, and Communication) in Brehmer´s design-logic hierarchy (Purpose, Function and Form). The results indicate that the method can be usefull to find requirements and design criteria for future C2-systems.

ABSTRACT. The capability to make quick, effective decisions under uncertainty is critical for Army mission success. One of the most promising ways to deliver and enhance this capability is through recommender systems, which predict needs in advance to prioritize networked information delivery. While many promising recommender methods exist that might fulfill this need, it is unclear how recommendations should be presented to C3I analysts. This is because recommender systems are typically studied in low-risk, high bandwidth domains, which has contributed to an increased focus on maximizing subjective user satisfaction (rather than objective success) while relying on copious amounts of global data. This is in contrast to C3I, which is often practiced under high-risk, low bandwidth (tactical networking) conditions. Since each decision made under such conditions can affect mission success and survival rate, analysts accordingly require a higher level of transparency and provenance for information systems. Moreover, since global uplinks are not always available, C3I analysts require the ability to process information locally in both time and space. Here, we describe two highly transparent, domain-oriented recommendation methodologies to support the needs of C3I analysts, depending on whether or not recommendations need to be calculated locally. Each method demonstrates its theoretical effectiveness on a popular recommender systems benchmark, but further evaluation is needed for the C3I domain.

ABSTRACT. The recent surge in cyber-physical systems (CPS) and Internet of things (IoT) (e.g., robots, industrial control systems) has resulted in increased network security challenges. In a distributed and semi-autonomous environment, network breaches must be detected prior to reaching the highly-valued targets or networked devices—this requires proactive adversarial modeling that is behavior or anomaly-based and capable of operating in a high-speed network environment. Using an intelligent agent architecture and machine learning, we propose a network intrusion detection system (NIDS) that is flow-based to produce alerts on malicious and/or anomalous traffic. With this proposed semi-supervised learning approach, we detect Botnet network traffic and distinguish it from the normal and background traffic in the IP flow datasets. We evaluate the prediction performance and computational resource utilization results for the flow-based NIDS algorithms and compare these results with signature-based NIDS that are reactive by design. With this approach, we show an improvement in detection accuracy and NIDS efficiency when compared with traditional signature-based NIDS and other probabilistic modeling approaches examined on these network traffic datasets. In addition, the model improvements reduce the burden on the human analysts to sift through NIDS alerts that are often riddled with false alarms.

ABSTRACT. The Social Terrain Modeling (STM) team at Army Research Laboratory applies a research framework built around the needs of Soldiers who must successfully navigate and utilize their environment. In the past, these environments have primarily consisted of physical terrains. However, the terrain of the future is not just physical, it is also social. Nonetheless, successfully navigating and utilizing this new social terrain to support the Army mission requires the same crucial capabilities: sensing of the terrain, measuring and modeling of the dynamics, evaluating the influence and communication between entities, and developing computational technologies to exploit the data and models. The paper overviews these four research directions of the STM team which form a progressive staircase leading to a better equipped and more capable Army of the future.

ABSTRACT. This paper examines the potential advantages of a cognitive sensor-to-effector loop (CStEL) , or ‘cognitive sensor-to-shooter loop’ (CStSL), concept over a non-cognitive one. The concept results from the author’s research on cognitive radars (CRs) and internet of intelligent things (IoITs). Also this paper contributes to his current research on cognitive radio network (CRN) in support of Defence Research and Development Canada (DRDC) science and technology (S&T) outlook to inform our organisation and the Canadian Department of National Defence about the potential operational impact of such technology advances. The envisaged CStEL is assumed to rely on intelligently internetworked cognitive components such as cognitive IoIT ecosystems which include sensors and effectors such as CR, cognitive electronic warfare (CEW), cognitive weapon (CW) such as cognitive missile or cognitively optimised directed energy weapon (CODEW) and CRN. According to literature surveyed, in some instances, CRs may offer improvement (gain) by one-order of magnitude in timeliness, accuracy and detection range which represent significant advantages to early adopters. Being at the cusp of practical specialized artificial intelligence in small devices which is critical to the CStEL concept, one may hypothesise that such CRs gain could be attained by CWs and by the cognitively supported decision process of CStEL, thus to expect that the overall system may offer one order of magnitude in improved interception rate success (positive outcomes of engagements). All of this can be seen as a result of significant progress of low power demand technologies advancing specialized and general artificial intelligence. This means less human intervention in the loop for local analytics which ensure more useful timely and actionable information to be shared.

ABSTRACT. With the emergence of advanced computational methods, enhanced insights into the social dynamics of complex operational environments, such as mega-cities, are now within reach. We contend the utilization of the standard Political, Military, Economic, Social, Information, Infrastructure, Physical Environment, and Time (PMESII-PT) framework to assess the operational environments within the Military Decision Making Process (MDMP) do not sufficiently address the subjectivity within the state space required to understand the target population or how their social landscapes interact with military operations. To close this gap, objective measures can be supplemented with new methods that utilize social-sense enabled informatics. Towards this end, we introduce the Pulse of the Population (POP) concept that augments the MDMP process. POP serves as a computational social science framework that enables on-going assessments of the population state space within varying military operational context. We postulate that conditions for desired end state in these complex landscapes are often shaped by social sensing demands which interact with a myriad of environmental factors. In this paper we describe how the POP concept would improve mission planning analysis, course of action development, and enable ongoing running estimates of the population state space within the area of operations.

ABSTRACT. Information extraction (IE) pipelines aim to point human decision makers toward relevant information, but beyond the accuracy of the pipeline itself, designing the presentation of the output of the pipeline for optimal human understanding should be a goal. This paper establishes a framework for testing comprehension of text documents with and without markup from an IE pipeline and reports the results of a behavioral experiment where an information extraction pipeline, instead of helping, seems to hurt both objective and subjective measures of performance. These results suggest further steps that can be taken toward developing more human-usable IE pipeline outputs.

ABSTRACT. In the early years of Artificial Intelligence (AI) research, scientists envisioned AI as embodied in a robot. Since then, the fields of AI and robotics diverged and mostly developed separately. Thanks to rapid developments and recent scientific breakthroughs in both fields, there are many upcoming applications where the integration of AI and robotics leads to great added value. In this article, we investigate the opportunities of integrated AI and robotics for the Defense domain. Based on recent scientific developments, we introduce the concept of a Non-human Intelligent Collaborator (NIC), an intelligent artificial team member with social and collaborative skills. We explore new applications of integrated AI and robotics across multiple domains, from agriculture to warehousing. The domains reveal a variety of problem characteristics, offering valuable lessons from the solutions presented, which can be translated to the Defense domain. We focus on the behavioural impact of deploying NIC-human teams on conducting future operations.

ABSTRACT. Western militaries continuously consider the future of armed conflict in order to better prepare for the challenges of tomorrow. This includes intelligence organizations monitoring trends to better understand threats in the near-term, and it also includes planners extrapolating trends into futures documents that present potential views of the future security environment. Defining the potential military missions for the future enables capability development toward a view of future challenges. The focus of this paper is on how emerging technologies such as Data Analytics, IoT-enabled technologies and Autonomous systems can provide new capability for combat service support for the future force, with specific focus on digital information solutions, and autonomous systems.

ABSTRACT. In addition to providing background information on US DoD Risk Management Framework (RMF) security controls and overlays and Agile C2 theory, this concept paper proposes research into the possibility of leveraging Agile C2 approaches to define notional system operational use cases and associated threat vectors to support the research and developing of an Agile C2-related security control overlays which would support the variety of missions, circumstances, and the collections of entities needed to meet these security challenges.

ABSTRACT. In 2016 a number of us proposed a Fifth Generation Headquarters Concept, by which Artificial Intelligence (AI) could be infused into military operational headquarters to enable it to achieve properties such as C2 Agility. In this paper, I advance the concept further by identifying what types of AI fit into the different functions of headquarters. I take the cue from the relationship between the two current realisations of AI in real technology to human modes of decision-making: Recommender Systems, based on statistical learning, and Expert Systems, based on branching rules and encoded logic. These map to the now well-known ‘two systems’ model of human decision-making popularised by Nobel Laureate Daniel Kahneman: system one, which represents intuitive thinking, and system two, describing human rationality. In turn, these two systems dominate, respectively, the modes of decision-making in what may be seen as the two ‘hemispheres’ of military headquarters: the J3 branch focused on current operations, and the J5 on longer term planning. Unlike J5, in J3 there is lack of time for elaborate rational business processes and thus staff adopt a mode of decision-making known as Recognition Priming. I suggest that a model by Henry Mintzberg for combining intuition-dominant ‘strategy formation’ with more analytical ‘strategic programming’ can become the basis for formulating the business processes of a Fifth Generation Headquarters. Finally, I discuss a number of AI driven mechanisms within such a construct that will support the property of C2 Agility in such a headquarters.