ARLINGTON, Va. – U.S. military researchers are asking industry for ways to encapsulate local and collective human experience into computational models of to give intelligence analysts and military commanders an “insider” view to enhance decision-making.
Officials of the U.S. Defense Advanced Research Projects Agency (DARPA) in Arlington, Va., issued a solicitation on Monday (HR001120S0035) for the Habitus project to create self-sustaining, adaptive, generalizable, and scalable ways to generate causal system models based on local knowledge to aid operational decision-making.
Understanding how to work with and influence local systems to support military operations is critical for decision-making, and is most challenging in undergoverned regions in which the systems themselves often change rapidly and are inaccessible to the U.S. military, DARPA researchers explain.
Humans develop cognitive models of systems of which they are a part, which include local customs, relations, and understanding. Unfortunately, these cognitive models often are largely inaccessible to outsiders.
This leads to problems like the failure to recognize how local burial practices may have influenced the spread of the Ebola virus in West Africa, and led to substantial errors in predicting of the disease’s spread in 2014.
U.S. early information operations in Iraq, moreover was ineffective because Iraqis distrusted radio and television broadcasts after years of Ba’athist propaganda. Efforts to secure a small town from ISIS created local turmoil when U.S. forces set up a perimeter that excluded the local glass factory and disrupted locals’ ability to get to work.
The primary challenge of the Habitus program focuses on creating of localized and collective human experience to give intelligence analysts and military commanders an “insider” view to enhance decision-making.
Big data methods often lack the ability to interpret data through a local lens. Exceptions tuned to one city often do not translate to other areas or populations. Instead, Habitus seeks a capability that is adaptive, scalable, self-sustaining, and detailed.
Projections must update to maintain accuracy, for example. Methodologies must be able to transfer to new regions, populations, and topics. This capability also must require few resources to maintain over time.
Habitus will capitalize on engagement with locals to build of collective local political, economic, and social experience to help decision makers understand not only what factors are important but also why and how these systems work from an insider’s view. In this way, military experts might be able to anticipate changes as events unfold, and help the military achieve its local goals.
The Habitus program will find ways to capture local knowledge, and make it available to intelligence analysts and military commanders. The program has two technical areas: model development, and engagement.
The program also will have three phases and last for 42 months. The first phase will build initial capabilities; the second phase will demonstrate scalability for in new regions; and the third phase perform an expended demonstration in a DARPA-specified region.
Companies interested should upload abstracts no later than 28 Feb. 2020, and full proposals no later than 23 April 2020 to the DARPA BAA Website at https://baa.darpa.mil/.
Email questions or concerns to Bart Russell, the DARPA Habitus program manager, at [email protected]. More information is online at https://beta.sam.gov/opp/2bfdcf9c688646ec8d01e1058f491dc4/view.
