Friday, January 16, 2015

Representation of Unmanned Systems in Naval Analytical Modeling and Simulation: What are we really simulating?

Editor's Note: This article is reprinted with permission from the Naval Postgraduate School's "CRUSER News.

By Professor Curtis Blais, faculty at the Naval Postgraduate School's Modeling, Virtual Environments and Simulation (MOVES) Institute. Contact: clblais(at) 

Combat models are used in major assessments such as Quadrennial Defense Reviews for Naval system acquisition and future force structure decisions. For several years, the Navy has been adding capabilities to the Synthetic Theater Operations Research Model (STORM) originally developed by the U.S. Air Force. Similarly, the Army and Marine Corps employ a specific analytical model called the Combined Arms Analysis Tool for the 21st Century (COMBATXXI) to evaluate major proposed changes in materiel and associated warfighting operations and tactics. The CRUSER Charter identifies numerous Naval initiatives for study and development of unmanned systems, such as the Unmanned Carrier Launched Airborne Surveillance and Strike (UCLASS) squadron, Large Diameter Unmanned Undersea Vehicles (LDUUVs), and an integrated Family of Robotic Systems to augment the capabilities of the Marine Air Ground Task Force (MAGTF) / Fleet.

Image Courtesy of NPS MOVES Institute
The Unmanned Systems Integrated Roadmap FY2013-2038 indicates the Presidential Budget for Fiscal Year 2014 was over four billion dollars (covering research, development, test, and evaluation, procurement, and operations and maintenance). With such current initiatives and high-valued expenditures occurring with respect to unmanned systems, there is concern that expected improvements to warfighter effectiveness, through tactics, techniques, or procedures, are not well supported by analytical processes and findings.

Initial investigation of models such as STORM and COMBATXXI that support studies for major decisions indicates that these simulations are largely deficient in representations of such emerging systems. Without such representations, it is not possible
to conduct studies investigating future force structures (e.g., 2020 and beyond) involving significant employment of unmanned systems. Instead, it appears that decisions are being made without an analytical basis that can show the benefits, limitations, and challenges (manpower, training, logistics,
combat service support, vulnerabilities, etc.) of introduction of such systems into the battlespace.
Starting in late 2014, we began investigating capabilities of these critical Naval analytical models to identify improvements needed in representations of unmanned system capabilities that can improve the scope and value of studies conducted using such tools. This is an initial effort to bring improved representations of unmanned systems into analytical environments, recognizing that it is part of a larger need to bring such representations into gaming environments for concept exploration, into constructive simulations for experimentation and mission planning, and into training environments for low-level (operator) to high-level (staff) skill development.

Interestingly, the initial research is raising a new thesis—that current analytical models actually possess, though unintentionally, a higher fidelity representation of autonomous systems than they do of human-operated systems! If this is true, users of current models must change their perspectives considerably. It is well recognized that a major challenge in modeling and simulation is representation of the human element in combat, reflecting human characteristics such as training, fatigue, unit cohesion, intuition, etc. The lack of such modeling extends to the operation of systems by humans, including the operation of robotic systems (teleoperated). In many respects, it may be argued that current models of the battlespace provide a reasonably accurate depiction of diverse land, air, sea autonomous systems interacting in the battlespace, while poorly representing the human element in the operation of warfare systems. How this change in perspective in understanding the capabilities and validity of current models will affect the modeling & simulation and analytical communities remains to be seen but clearly needs further study. A key issue becomes determining how to better distinguish humans and human-operated systems from autonomous systems so that the models can more correctly represent all of these systems, and their interactions, in the battlespace.

Monday, January 12, 2015

Largest Autonomous Underwater Vehicle Swarm

Researchers at Austria's University of Graz have demonstrated the largest collection of swarming autonomous underwater vehicles with their Collective Cognitive Robots (CoCoRo) project.  A total of 41 autonomous underwater vehicles (AUVs) were assembled for recent swarm testing at the University's Artificial Life Lab. Though funded by the European Union's Seventh Framework Programme for Research (FP7) with the intention of developing civilian innovations for environmental monitoring and research, CoCoRo has implications for future military unmanned underwater vehicle swarm activity. 

Under development since 2011, CoCoRo's swarm demonstration consists of three types of robots: Jeff is an agile fish-like robot with various pressure and magnetic sensors for obstacle detection, avoidance, and navigation.  The swarm also featured 20 saucer-shaped Lily robots that randomly search for objects while communicating with each other using blue-LED lights.  The final robot is a semi-submersible catamaran base station which serves as a platform for the vehicles to autonomously dock allowing the swarm to communicate its location (via GPS) and activities with humans as well as to other base stations.  Eventually, using this method, swarms in multiple geographic areas could coordinate search areas with one another. A dock could also provide a future means to recharge the AUVs and transport them from one location to another.
Bio-inspired algorithms enable the swarm to work together to locate magnetic targets and aggregate around them.  On a larger scale, this behavior has viability for naval unmanned underwater vehicles that could be used in underwater surveys, search and recovery, or mine counter-measures operations