|
A Cognitive Architecture for Naval Mine Countermeasures (MCM)
Navy SBIR 2012.1 - Topic N121-082 ONR - Ms. Tracy Frost - [email protected] Opens: December 12, 2011 - Closes: January 11, 2012 N121-082 TITLE: A Cognitive Architecture for Naval Mine Countermeasures (MCM) TECHNOLOGY AREAS: Information Systems, Battlespace ACQUISITION PROGRAM: Mine warfare Environmental Decision Aid Library (MEDAL), PMS-495, PEO-LCS OBJECTIVE: Develop a cognitive architecture for naval MCM planning and re-planning that will significantly reduce the time required to re-plan and produce plans with a measurable decrease in risk to the warfighter via better management of uncertainty. DESCRIPTION: MCM missions involve both objective and subjective decisions by military commanders that balance risk versus time and are based on many factors all with non-trivial uncertainty. A staff then translates this commander's intent into actionable orders using a combination of judgment and automated algorithms. This process is manpower intensive, poorly quantifies/articulates uncertainty, and is not well-suited to real-time re-planning. Cognitive architectures are designed to address these deficiencies in general yet have never been investigated for MCM. The aspects of MCM to be investigated include mission planning for highly dynamic and uncertain environments and constraints, real-time reasoning on both mission progress and how/when to re-plan, and fluidly interfacing with human operators to enable better decisions as well as instill trust in the underlying algorithms and computer-generated solutions. The overall goals of this effort are to: reduce mission re-planning time; better articulate uncertainty, context, and constraints to human operators; and, provide a transparent interpretation between (mission-level) commander's intent and (unit-level) actionable orders. A successful development will have two paths for transition; the first is to transition directly to the fleet's tactical decision aid used for MCM planning and evaluation, the Mine Warfare and Environmental Decision Aids Library (MEDAL). The second path is to transition to MEDAL via an FY13-start ONR Enabling Capability program that will modernize and expand the functionality of these tactical decision aids and their underlying algorithms. State-of-the-Art: The maturity, effectiveness, and employment of cognitive architectures have been expanding significantly in recent years. Their benefits over alternative approaches include representations that are more easily/naturally understood by humans and the ability to reason/operate on complex problems with uncertain and incomplete information. However, cognitive architectures have yet to be applied to MCM even though their benefits are well matched to MCM's current deficiencies. References are provided for general information; however, there is no preferred or prescribed architectural solution associated with this solicitation. PHASE I: Investigate the processes for MCM planning, evaluation, and re-planning; identify the most appropriate, highest payoff areas; and develop an initial high-level architecture design. The option period will initiate a detailed architecture design and begin modeling key elements. The government will provide relevant MCM information to Phase I recipients (after award) and relevant MCM databases to those awarded the option--detailed knowledge of MCM is not required prior to Phase I award and need not be demonstrated in the proposal. All work in this phase will be UNCLASSIFIED. PHASE II: Develop a prototype cognitive architecture to be fully functional with government-provided databases, simulation capabilities, and interface descriptions. Initial development will be standalone with a goal of integration into the program of record. The option period will be used to expand the scope of the MCM problem addressed by the architecture and to integrate into the program of record. Work in this phase may be done at the UNCLASSIFIED level; however, the ability to handle SECRET databases would add significant flexibility. PHASE III: Extend the architecture to operate effectively, be robust, and be fault tolerant to a full spectrum of government-provided data, constraints, and planning conditions. This will involve significant coordination with the government laboratory to fully integrate and test in the program of record. PRIVATE SECTOR COMMERCIAL POTENTIAL/DUAL-USE APPLICATIONS: This capability is applicable to undersea survey (e.g., petrochemical, geological, etc.) as well as non-maritime search and survey. It is also applicable to the general problem of supervision of multiple robotic or unmanned systems. REFERENCES: [2] Cassimatis, N. L., Murugesan, A., & Bignoli, P. (2009). Inference with Relational Theories over Infinite Domains. In Proceedings of the FLAIRS-09 Conference, Florida, USA, pp. 21-26. [3] Pearson, D. J., Laird, J. E. (2005). Incremental Learning of Procedural Planning Knowledge in Challenging Environments, Computational Intelligence, 21:4, 414-439. KEYWORDS: cognitive architecture; human-system interface; mine countermeasures; mission planning; re-planning; uncertainty
|