Extending Cognitive Assistance with AI Courses of Action

Thumbnail Image
Runde, Sharon M.
Godin, Arkady A.
decision making
decision support
cognitive assistant
artificial intelligence
fire support
Date of Issue
Monterey, California: Naval Postgraduate School
The objectives of this study is to research and assess the initial stages of the evolution of Human-Machine Teaming (HMT) mission workflows which is focused on transitioning of automation tasks from humans to machines using a technique to digitize mission workflows. Also, study the advanced stage(s) of the evolution of HMT to include Courses-of-Action (COA) in Wargaming and how decision-making (DM) AI functions play what role natural language processing (NLP) plays. In addition, this study will explore the viability of NLP in HMT peer-to-peer COAs generation. Finally, this study will leverage complex Joint Naval Force EABO scenario (UNCLASS) designed by MCWL to explore NLP and distributed agents managing the decision making of operators using various modes of HMT interface of AI run-time execution agents thereby enriching digital workflows. The research questions that will be address will include: 1) What is the best approach for a cognitive assistant to learn mission workflows so that recommendations can be made to a human operator?, 2) How can cognitive assistants switch between modes of automatic, advisory, or monitoring?, 3) What are the key parameters for switching?, 4) How does the CA learn to switch to make appropriate recommendations?, 4) What is the cognitive intersection between domain specific environment awareness and situation awareness?, and 5) What happens when a target switches context? The methodology will use quantitative research methods. The methodology for this study will be based on SME input to gain an understanding of mission workflows and tasks, MCWL-developed Joint Force EABO scenario leveraged for a case study and collaboration with the Wargaming Center in Quantico, VA. Based on a scenario, the independent variables will be the inputs into the cognitive assistant. The dependent variable(s) are the output of the system such as if the system recommends the role of automatic, advisory, or monitoring. The plan for this study is to leverage a complex joint Naval Force EABO scenario in studying a role of enrichment digitization of the workflows including utilization of scenario-driven HMT modes and sub-modes; review digital workflows from Master Thesis: "Fire Support Coordination Cognitive Assistant", USMC Capt. Benjamin Herbold, NPS, Graduation Year: June 2020; gain understanding of wargaming COA Digital Mission Command Joint Forces hypergame; develop expertise on modes of Human-Machine Teaming control and their sub-modes of automatic, advisory, and monitoring; study evolution from a single "interactive" mode of HMT proposed for the Fire Support Coordination Digital Workflows to the planning phase in Fire Support Coordination; study NLP and associated theories as a framework to situate the research; and coordinate with other entities such as MIT LL, DARPA, BAE, USMC AI COI, MCWL, and ONR.
Technical Report
NPS NRP Technical Report
Information Sciences (IS)
Other Units
Naval Research Program (NRP)
NPS Report Number
HQMC Plans, Policies & Operations (PP&O)
This research is supported by funding from the Naval Postgraduate School, Naval Research Program (PE 0605853N/2098).
Chief of Naval Operations (CNO)
Distribution Statement
Approved for public release. Distribution is unlimited. 
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.