You are here

Art + Spatial Geospatial Innovation


OUSD (R&E) MODERNIZATION PRIORITY: AI/ML TECHNOLOGY AREA(S): Causality and Inference Discovery, Machine Learning, Complex and Dynamic Graph Theory, Information Systems OBJECTIVE: Applicants are to propose methodologies to analyze and describe operational environment (OE) complexity in terms of the above definition through development pathways to elevate the cognitive ability of machine learning (ML) and artificial intelligence (AI), and convergence of cognitive diversity into technology applications. The three levels cognitive ability is defined as follows: the lowest as ‘seeing and observing’ - detection of regularities in environments; the next level up as ‘doing’ - predicting effects of deliberate alterations to produce a desired outcome; the highest as ‘knowing’ - understanding the (causal inference) of why something works and what to do when it does not. Cognitive diversity is defined as different manners of thought, generating ideas, problem-solving methods and perspectives. DESCRIPTION: Military commanders and key leaders are seeking and continually ask their staffs for the operational ‘so what?’ Conflict, social disruption, disease, strain on resources, climate change, and economic instability are formed upon obscured, complex and dynamic factors and make the identification of meaningful and actionable ‘so what’s’ extremely difficult. Understanding such complexity requires in-depth cognitive ability and cognitive diversity. Leaders, planning staff, analysts, operators must possess extensive expertise and pour through enormous data sets and information to understand what the ‘so what’ is and know how to present the ‘so what’ in a manner that commanders and leaders can make decisions from. Barriers to better understand the OE and plan for operations include: a lack of qualitative analytic outputs that provide multiple perspectives understanding; a lack of analytic capabilities to describe OE conditions in terms of system’s behavior and causation; and the inability to accurately and dynamically describe the attributes of edges within the system. To overcome these barriers, the Army should not simply build better “analytic mouse traps.” Current technology applications focus on ‘big data’ with limited means to provide meaningful interpretations and expressions of causation. Although ‘big data’ is in fashion, it is no panacea. It is neither an end state nor is it a way for gaining higher levels of cognitive capabilities. Rather, meaningful outputs are comprised of contextual descriptions of OE conditions and the progress towards or regress from specified objectives. This research proposal seeks to support the generation of means for creating technologic methodologies to make such determinations. Enablement of cognitive ability and cognitive diversity offers pathways of discovery beyond identification of system nodes and their attributes. The vision behind this form of research is the generation of greater understanding of system behaviors (in context of operational variables) through exploration of system edges. The research seeks to support the development of methodologies for collectively identifying, examining, and systematically integrating edge attributes (e.g. relationship strategies, motivations, and expected outcomes) into a collaborative analytic platform for operational and strategic staffs to determine patterns of system behaviors within OEs. Central to this development is a novel integration of artist into the design and development process for this effort. As critical as the development of an analysis capability that can capture complex and dynamic system behavior that reveals actionable levers of control within that system with a focus on edge attributes, is the development of a visualization capability that both captures the richness and complexity of the system behavior in the OE, while, making it readily understandable what the levers of control are in the multi-domain environment and providing the ‘why’ and ‘how’ of these levers by providing a human understandable causal links to these levers. The key here is both to reduce the cognitive burden and training requirement for the tool, while providing the warfighter a capability that answers the critical “so what” question for the commander. PHASE I: The objective of this phase will be to accomplish two primary tasks: a) develop technical approach that is capable of ingesting and analyzing a combination of warfighter gathered, open source and publicly available information and data to support the situational understanding required to identify the gap between a commander’s current state in the mission space and their goal state, along multiple interrelated and interacting lines of effort. The second task, b) is to concurrently collaborate with artists and other researchers skilled in innovative interpretations and novel visualizations to collectively develop analytical visual outputs for this capability that is intuitive, minimizes cognitive burden and training, while providing actionable insights to the commander in a complex and dynamic environment. The goal would be to create a study that would include and assessment of alternative approaches, along with the risks of each approach and risk mitigation strategies for each alternative. Although not required, a simple prototype that demonstrates the offeror’s best of breed approach for Phase 2, with a focus on novel visualization to reduce cognitive burden would be beneficial. PHASE II: The objective of phase 2 would be to create a fully functional prototype of the capability design to support a small selection of use cases for operational warfighters that will be significantly impacted by the human element of the operational environment. By providing an operational use case to focus this effort, we provide a more realistic opportunity for the offeror to be able to deliver a practical capability that will meet the needs of operational users while also being able to demonstrate the power of this approach to analyze edge attributes and demonstrate levers of control or actionable information to the warfighter in a human explainable way. PHASE III DUAL USE APPLICATIONS: The goal of this topic is to upgrade the cognitive ability of AI/ML when scanning the information environment. More intelligent, context-aware AI is in-demand for multiple industries. Therefore, in phase 3, the goal would be to expand this development into non-military domains that would include logistics, marketing campaigns, emergency response management and on-line information/disinformation campaigns, just to name a few non-military examples. REFERENCES: Page, Scott E. 2018. The Model Thinker: What You Need to Know to Make Data Work for You. Basic Books, Inc. New York, NY. Pearl, Judea, and Dana Mackenzie. 2019. The Book of Why: The New Science of Cause and Effect. Harlow, England: Penguin Books. “Cognitive diversity: The diversity your company isn't thinking about.” September 6, 2021. Accessed from:,solving%20methods%20and%20mental%20perspectives. KEYWORDS: Geospatial; data analysis; visualization; risk mitigation
US Flag An Official Website of the United States Government