Papers by Georgiy Levchuk
Cognitive Particles is a step toward realizing the ability to develop autonomous components (part... more Cognitive Particles is a step toward realizing the ability to develop autonomous components (particles) that are capable of coordinating to come together and form a desired object. Automated shape assembly and disassembly from a collection of particles would allow incredible resource availability and flexibility in domains ranging from the highly specialized and time-critical (soldiers in the field or medical technicians in the operating room) to the more mundane. In this work, we first developed theoretical concepts supporting the development and coordination of autonomous shape-forming particles. Building on this foundation, we constructed a 3-dimenstional particle simulation testbed to enable experimentation regarding S
Using Government drawings, specifications, or other data included in this document for any purpos... more Using Government drawings, specifications, or other data included in this document for any purpose other than Government procurement does not in any way obligate the U.S. Government. The fact that the Government formulated or supplied the drawings, specifications, or other data does not license the holder or any other person or corporation; or convey any rights or permission to manufacture, use, or sell any patented invention that may relate to them. This report was cleared for public release by the 88 th ABW, Wright-Patterson AFB Public Affairs Office and is available to the general public, including foreign nationals. Copies may be obtained from the Defense Technical Information Center (DTIC) (http://www.dtic.mil).
Intelligence Preparation of the Battlefield (IPB) provides critical support for military planning... more Intelligence Preparation of the Battlefield (IPB) provides critical support for military planning and decision making. In both offensive and defensive operations, the IPB process gives a decision maker information about the enemy, including potential courses of action, as well as information about the environment in which he is working. Physical structures that can support repetitive crimes (such as IED supply chains or illegal drug trafficking) provide important information for C2 planning. Activities of interest must be situated somewhere, and the physical structures present in any given location change slowly. Knowledge of those structures and their capabilities therefore provide an effective lens through which to view activities, and therefore an effective means for attacking the problem. In this paper we discuss the Facility Identification via Networks with Adaptive Links (FINAL) technology that Aptima is developing to find facilities associated with adversarial actions and dis...
Lecture Notes in Computer Science, 2009
Model-driven simulation can make the design and delivery of instruction more efficient and effect... more Model-driven simulation can make the design and delivery of instruction more efficient and effective. We describe two computational models that support both the design and delivery of instruction. BEST (the Benchmarked Experiential System for Training) can guide experts through the space of domain problems during the knowledge engineering phase of instructional design; it can guide trainees through the space of training objectives during instruction. PRESTO (Pedagogically Relevant Engineering of Scenarios for Training Objectives) builds scenarios on the fly to elicit the knowledge of experts during instructional design, and to satisfy the instructional objectives of trainees.
As the scale of automated and unmanned forces increases, a crucial challenge of controlling, coor... more As the scale of automated and unmanned forces increases, a crucial challenge of controlling, coordinating, and synchronizing the operations of individual force components must be solved. In asymmetric future environments, including urban security operations or disaster relief missions, the traditional hierarchical command and control (C2) organizational structures would not provide enough flexibility needed to adapt the tactical operations. To enable agile functioning of the force, instead of such centrally-controlled operations, researchers recently explored the fully distributed control strategies. However, fully distributed operations are only possible for simple team missions, and are not effective in executing the plans that have complex resource, temporal, and relational dependencies. Our previous research focused on designing hybrid organizational command, control, communication, and information structures, which combine benefits of hierarchical and heterarchical organizational principles to enable agile operations. In this paper, we describe the application of hybrid C2 design to the domain of automated shape assembly, which we performed in the project called Cognitive Particles sponsored by DARPA, Defense Sciences Office. The objective of this research was to develop the processes and control architecture for the automated particle-based shape formation for a large variety of shape structures and under significant uncertainties in the construction process. The roles of commanders are fulfilled by "active particles", and the roles of force units are fulfilled by "reactive particles". The particles coordinate to execute the overall plan based on specified shape objectives; these processes are similar to command, control, and coordination requirements of automated and semi-automated forces. Current particle formation frameworks are based on the rigid skeleton assemblies, which can only handle a fixed set of predefined shapes, and cannot adapt the construction to the action uncertainty or changes in the environment. As such, the particle-based shape assembly provided a perfect case study to test the concepts of the agile organizational structures and processes. Motivation Planning and Controlling Semi-Automated Forces Traditional planning, organizational design, and tactical execution processes of U.S. military depended on having relatively complete knowledge of the threat (e.g., composition of enemy forces, doctrine, likely operational and tactical situations and geographic conditions). However, both conventional adversaries and asymmetric threats confronted in the Current Operational Environment (COE) can no longer be fully engaged using conventional approaches and organizations. They require more facile, dynamic organizational structures that enable agile and precise operational and tactical actions. As a response to volatile environments, organizations struggle to balance stability against flexibility, specialization against generalization, and centralization against decentralization (Alstyne, 1997).
IEEE Transactions on Communications, 2004
A fast optimal algorithm based on the branch and bound method, coupled with an iterative lower bo... more A fast optimal algorithm based on the branch and bound method, coupled with an iterative lower bound update, is proposed for the joint detection of binary symbols of K users in a synchronous correlated waveform multiple-access (CWMA) channel with Gaussian noise. Although the group optimal detection problem is generally NP hard, the proposed method can significantly decrease the average computational cost. A fast "any-time" sub-optimal algorithm is also available by simply picking the "Current-Best" solution in the Branch and Bound method. Theoretical results are given on the computational complexity and the performance of the "Current-Best" sub-optimal solution. Although the performance of sub-optimal algorithms are affected by the detection order, an order algorithm is proposed and the optimal detection order is shown to be identical to all the proposed sub-optimal algorithms. Simulation results are presented to verify the theoretical analysis. In the situation when only a small number of users are correlated, sub-optimal algorithm outperforms the decision feedback method significantly and the computational cost can be even less than that of the conventional method.
... 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Aptima Inc,12 Gill Street Suite 1400,Wobur... more ... 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) Aptima Inc,12 Gill Street Suite 1400,Woburn,MA,01801 ... GEVA: 2 SAR GCDL, GSML: 1 STRK GSAM: 2 TLAM (from 2 different platforms) GSA3: 2 STRK (1 F18S) GSA6: 2 TLAM (from 2 different platforms) GRGF ...
This paper describes a quantitative Team Optimal Design (TOD) methodology and its application to ... more This paper describes a quantitative Team Optimal Design (TOD) methodology and its application to the design of optimized manning for E-10 Multi-sensor Command and Control Aircraft. The E-10 (USAF, 2002) is intended to consolidate command and control (C2), battle management (BM), intelligence, surveillance and reconnaissance (ISR), and selected information warfare (IW) functions and eventually replace elements of the current C2ISR force mix. Our TOD design approach allows specification of team configurations (number and roles of operators) for efficiently operating E-10 to execute mission scenarios of various contingencies. The objective is to maximize the speed of mission execution while balancing the workload among team members, provided the decision-making, expertise, workload threshold and organization cost constraints are satisfied. In order to analyze and simulate the operations of E-10, we conducted mission decomposition to define functional responsibilities for the E-10 within the context of an operational mission. Working with subject matter experts, we developed functional process flows of E-10 by decomposing each stage into representative functions. Each of these functions is in turn decomposed into high level task responsibilities, defining a directed graph of tasks with precedence/information flow constraints.
What is Heterarchy? A Heterarchy* is a form of organization resembling a network, where authority... more What is Heterarchy? A Heterarchy* is a form of organization resembling a network, where authority is determined by knowledge and function A decision heterarchy is made up of a group of humans with common purpose. The group is organized horizontally with all individuals sharing equal authority and equal responsibility. *The term heterarchy was introduced into science more than half a century ago by the neurophysiologist and cybernetician Warren St. McCulloch in his study "A Heterarchy of Values Determined by the Topology of Nervous Nets" ® UConn 3 Why Study Heterarchies? Enabling Capabilities Modular Forces Networking Research Needs
This paper presents an optimization-based agent-driven Distributed Dynamic Decision-making (DDD) ... more This paper presents an optimization-based agent-driven Distributed Dynamic Decision-making (DDD) simulation model to evaluate the SPEYES (Security and Patrolling Enablers Yielding Effective SASO-Support and Stability Operations) system. The key challenge is to quantify the force multiplying effect of SPEYES technologies, which span sensing, situation awareness/command and control (SA/C 2), and shaping components. The performance improvements were measured in terms of timeliness, effectiveness, and efficiency of operations. The behaviors of optimization-based agents were first calibrated to those of human-in-the-loop simulations. The agent-driven simulation results indicated that integration of SPEYES sensing, SA/C 2 , and shaping technologies provided significant performance improvements to the force across all measures. Even at 50%-reduced force, the SPEYES system maintained significant performance improvements over regular operations with a full force and without SPEYES, thus confirming the force multiplier effect of SPEYES technologies. The findings are confirmed by human-in-the-loop simulations. I. INTRODUCTION A. SPEYES System Overview This paper discusses an agent-driven DDD simulation model to assess the SPEYES system. Historically, rapid post-conflict stability has been attained through high-troop densities. A key motivating theme leading to the SPEYES effort is the challenge of trying to enhance the SASO effectiveness with a limited number of forces through various force-multiplying technologies. The force multiplying technologies include: Sensing-low-cost, easily-emplaced, camouflaged sensors (video, acoustic, Infrared) to provide urban situation awareness; Situation Awareness/Command and Control (SA/C 2)-multi-echelon software tools tailored for distributed operations (planning, dynamic resource management, simulation, mission rehearsal); and Shaping-non-lethal, and explosives ordinance disposal (EOD) shaping tools, for example, to diffuse adversaries, crowds, and improved explosive devices) [1]. Figure 1 illustrates notionally how 'security' coverage could be expanded with the introduction of various technologies, e.g., tracer RFIDs, non-lethal weapons, smart scheduling, multi-spectral cameras, swarming unmanned vehicles. The force-multiplying enablers increase the effectiveness of forces for SASO by means of better threat recognition, increased situational awareness and intelligence of the post-conflict environment, and improved decentralized operations. The SPEYES sensing technologies will increase the speed and ability of the forces to gather intelligence by the combined use and smart placement of visual, acoustic, spectral, and infrared sensors. The information gathered from the sensor observations allows troops and commanders to gain better situational awareness, and facilitates more efficient scheduling and patrolling, including the utilization of innovative shaping technologies [6]. Better and faster observations, combined with more
This paper presents a software environment for adaptive organizational design, with focus on synt... more This paper presents a software environment for adaptive organizational design, with focus on synthesizing Joint Task Force (JTF) C2 architectures subject to organizational constraints (e.g., the availability of resources and/or DMs, the distribution of DMs' expertise, etc.). Currently, the design environment includes software modules for: (1) mission modeling to extract task dependency graphs; (2) mission planning to allocate resources to tasks; (3) hierarchical clustering algorithms for grouping resources into decision-maker (DM) nodes, and (4) building an organizational hierarchy. In addition, basic modules for dynamic adaptation of organizational strategies and structures in the face of changing mission environment and/or resources are being added. The organizational design environment presented in this paper enables an analyst to synthesize robust organizational structures and evaluate their performances. The software tool allows an analyst to decompose the process of organizational design into a sequence of stages and visualize the design process. The software also allows a user to input parameters and constraints in a natural way at various stages of the design process, making it possible to design organizational structures with desired attributes (e.g., speed of command, workload, team coordination).
To counteract the enemy organization, knowledge of the principles under which this organization o... more To counteract the enemy organization, knowledge of the principles under which this organization operates is required. This knowledge provides the ability to detect and predict the activities of the enemy and to select the appropriate counter-actions. Certain counter-actions require additional knowledge about enemy organization and processes-ranging from the specifics of organizational command, control, communication and information distribution (C3I) structures to the responsibility delegation and goals at the most important enemy nodes. Our paper proposes to solve the problem of identifying the enemy organization and activities via the NetSTAR system employing a hybrid multi-phase model-based structure and process identification approach. The basis for NetSTAR is an innovative methodology that integrates a social network model of coordination, a meta-task model of enemy goals, and a Hidden-Markov Model (HMM) of enemy activities to detect subgroups engaged in coordinated activities. This model enables the computation of the likelihood of the hypothesized organizational structure and processes given the observed behavior, and allows designing effective dynamic counter-action strategies via Partially Observable Markov Decision Processes (POMDP) modeling.
Public reporting burden for the collection of information is estimated to average 1 hour per resp... more Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number.
Goal management is the process of recognizing or inferring goals of individual team members; aban... more Goal management is the process of recognizing or inferring goals of individual team members; abandoning goals that are no longer relevant; identifying and resolving conflicts among goals; and prioritizing goals consistently for optimal team collaboration and effective operations. A Markov decision process (MDP) approach is employed to maximize the probability of achieving the primary goals (a subset of all goals). We seek to address the computational adequacy of an MDP as a planning model by introducing novel problem domain-specific heuristic evaluation functions (HEF) to aid the search process. We employ the optimal AO* search and two suboptimal greedy search algorithms to solve the MDP problem. A comparison of these algorithms to the dynamic programming algorithm shows that computational complexity can be reduced substantially. In addition, we recognize that embedded in the MDP solution, there are a number of different action sequences by which a team's goals can be realized. That is, in achieving the aforementioned optimality criterion, we identify alternate sequences for accomplishing the primary goals.
In this paper, we present techniques for modeling the decision-making processes of a team of synt... more In this paper, we present techniques for modeling the decision-making processes of a team of synthetic agents operating in task selection and resource allocation settings within the third generation distributed dynamic decision-making (DDD-III) paradigm. The DDD-III simulator provides a controllable, multi-player, multi-platform organizational environment. The paper provides two major contributions. First, motivated by the need for a network of intelligent agents within C2 experimental settings, a brief overview of modeling techniques for the design of a network of collaborating agents is provided. Second, techniques for modeling the decision-making processes of synthetic agents in task selection and resource allocation settings are presented. In the proposed framework, the decision-making processes of a network of intelligent agents are addressed via limited look-ahead, auctionbased scheduling and resource allocation algorithms from the phase I of the three-phase organizational design process. Preliminary results of operationalizing the DDD-based multi-agent-network paradigm are presented in two different mission scenarios. A coordination-free scenario illustrates the basic structures of the intelligent agent design, in terms of stimulus-hypothesis-option-response (SHOR) model-based three-stage decision-making process. The second example, derived from the A2C2 Experiment 8, highlights the potential of utilizing the agent framework in C2 experiments.
Uploads
Papers by Georgiy Levchuk