CHI

Unmanned Systems Training and Analysis Projects

Fundamental Interactions Generating Heterogeneous Teams (FIGHT-R)

CHI Systems is collaborating with DARPA to design the Fundamental Interactions Generating Heterogeneous Teams (FIGHT-R). Projects include robots architecture to create more effective heterogeneous teams of multiple robots and humans working cooperatively to perform specialized tactical maneuvers involved in close-quarters combat (CQC). Our specific focus is on those aspects of robot control software development, robotic system design, and human-robot interaction (HRI) R&D essential for robots to work with human team members in small units on tasks like clearing a room. This focus enables dynamic teamwork and role coordination for all members of heterogeneous teams.

The following are key features being developed in the FIGHT-R project:

Role mechanisms which allow the robot to understand its own and teammates’ rights, behaviors, and obligations.

Shared mental model, encapsulation of the domain in a common mission “story” which will allow the robot to understand how its present maneuver fits into the larger operation.

Symbolic abstraction of sensor data which allows the robotic teammate to understand, process, and communicate information about its environment in a manner similar to human teammates.

Logic engine for robotic behaviors, to provide movement and sensor reporting capabilities.

An effective, efficient human-robotic interface, for direction of the robotic teammate.


Motive and Affect-based Robotic Control (MARC)

CHI Systems is supporting the Office of Naval Research (ONR) in addressing two core challenges to the wider and more effective use of robotic and uninhabited vehicles (UxVs). They include the need for more robust and understandable autonomy and the need for simplified, intuitive, and more learnable human control mechanisms. We are employing cognitive models of personality and emotion to exploit the deep-structural human capabilities to communicate and give directions in terms of goals, motives, and emotions, thus, making human-to-UxV communication more natural and intuitive. Our novel approach seeks to map ordinary personality and emotion language to the kind of commands a UxV controller might use in communicating with unmanned systems.

Copyright © 2010 CHI Systems Inc. All rights Reserved | Terms of Use | Visit OSI Geospatial Corporate