|
The following summarizes past and currently funded projects for Cognitive
Systems research and development at Sandia National Laboratories
Externally
Funded Projects
DARPA Augmented
Cognition
Project Title:Engineering a Transformation
of Human-Machine Interaction to an Augmented Cognitive Relationship
Sandia work in support of the DARPA
Augmented Cognition program commenced in April of 2002. The
objective for the DARPA program is to "extend, by an order
of magnitude or more, the information management capacity of the
human-computer warfighter." Initially, emphasis has been placed
on detection of an operator's cognitive state so that systems may
adapt accordingly (e.g., adjust information throughput to the operator
in response to workload). Work conducted by Sandia focuses on development
of technologies to infer an operator's ongoing cognitive processes,
with specific emphasis on detecting discrepancies between machine
state and an operator’s ongoing interpretation of events.
This work is being conducted in collaboration with the Air Force
Research Laboratory and Notre Dame University.
Contact: Chris Forsythe, jcforsy@sandia.gov
Sandia
Internally Funded Projects
Augmented Cognition
Grand Challenge
Project Title: Cognition-Driven Data Analyst: Augmenting Analysts’,
Engineers’ and Scientists’ Capacity to Interpret Diverse
Data
This project focuses on development of technology solutions that
augment the capacity of analysts, engineers and scientists to detect
and correctly interpret meaningful patterns based on large volumes
of real-time and archival data derived from diverse sources.
There are three elements
to our technical approach. First, cognitive models are developed
that possess knowledge consistent with specific individuals (e.g.,
different domain experts, representatives of different cultural
groups). Here,
we build upon a proven capability to computationally model the cognitive
processes whereby an individual recognizes meaningful patterns (i.e.,
situations, contexts, storylines) based on combinations of environmental
cues. Second, a cognitive model is developed that captures the knowledge
of a specific user. Third, various mechanisms engage the user in
interactions that promote the ability of the user to recognize and
incorporate the perspective of other disciplines, cultures or interests
into their own thinking. The solution offered here not only engages
the user in the decision making process, as opposed to automating
the process, but interacts with the user in a manner that is tailored
to a user’s unique knowledge and experience.
Contact: John Wagner, jswagne@sandia.gov
Knowledge Capture
Project Title: Extensible Knowledge-Based Agents for Simulation
An essential step in developing agent-based simulations for any
application involves the representation of knowledge for the application
domain. This project was undertaken with the objective to expand
existing capabilities for human modeling and simulation to facilitate
their application to new domain problems. Specifically, this project
has emphasized the development of techniques for knowledge elicitation
and modeling to support creation of individualized models of naturalistic
decision making processes.
Contact: Ann Speed, aespeed@sandia.gov
Episodic Memory
Project Title: Endowment of Simulator Agents with Human-Like
Episodic Memory
Human information processing relies heavily on representations of
experiences that occur during the course of a lifetime. These memories
are not simple recordings that can be played back at recall, but
meaningful representations of experiences. They also provide the
common ground for day-to-day interactions with other people. The
objective is to develop a capability that would endow a machine
with similar ability to store and retrieve experiences. This capability
would allow the machine to reason at a symbolic level, based on
past success and failure. It would also allow common ground in the
interactions between humans and machines, and between multiple machines
(e.g., one machine may download its experiences to another machine).
Contact: Ann Speed, aespeed@sandia.gov
Naturalistic
Decision Making Cognitive Model
Project Title: Comprehensive Representation of Human Decision
Processes
While there has been substantial progress in modeling and simulation
of human agents, practical application is limited due to questions
concerning the realism of agent behavior. This project provided
a first step toward a framework for highly realistic human emulation.
Initially, emphasis was placed on computationally representing human
Naturalistic Decision Making. The conceptual approach resulting
from this work then served as a foundation for development of a
framework for the comprehensive representation of decision processes.
This framework utilizes a two-tiered architecture in which an underlying
physiological model serves as the engine for a psychological model.
Knowledge and cognitive processes are represented within the psychological
model, whereas the physiological model provides the basis for incorporating
organic factors (e.g., arousal, emotion, etc.).
Contact: Chris Forsythe, jcforsy@sandia.gov
|
