This project is being
conducted by Sandia National Laboratories in support of the DARPA
Augmented Cognition program. Work 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. Project
Description
In Phase 1 of DARPA's Augmented Cognition program, Sandia has developed
a capability referred to as Discrepancy Detection. The objective
is to enable a machine to detect when an operator has committed
a situation mode error (i.e., operator believes the system is in
one state and behaves accordingly, when the system is actually in
a different state).
As illustrated in the
Figure below, we first develop a detailed cognitive model of an
operator's knowledge of a task. Currently, our development has focused
on an AWACS operator. The machine then uses the operator's cognitive
model to infer its own state. Simultaneously, the machine uses data
from the operator to infer the operator's beliefs concerning the
state of the system. This data currently includes actions taken
by the operator and eye-tracking to identify cues that have received
attention from the operator. By comparing the machine's recognition
of its own state with the operator's behavior, the machine may detect
conditions in which a discrepancy exists between the actual state
of the machine and the operator's beliefs concerning the state of
the machine.

Our immediate plans involve
integration of physiological indices with our capability for Discrepancy
Detection. We believe that real-time EEG measures will allow us
to enhance our capability to infer cognitive processes, while output
from our cognitive model will also enhance the ability to infer
cognitive workload based on physiological measures. Thus, the two
capabilities should combine synergistically to enhance the performance
of each.
Collaboration
This project is being conducted in conjunction with three collaborating
organizations:
Air Force
Research Laboratory – developing a multi-modal
human-computer interface to support Discrepancy Detection within
the context of the AWACS operator task interface.
Notre Dame University – conducting basic
research to gain an understanding of the mechanisms whereby Discrepancy
Detection occurs within the context of human-to-human interactions.
Project Documents
Presentation July 2002 DARPA AugCog PIs Meeting – SNL [PDF
file]
Contact Chris Forsythe (jcforsy@sandia.gov)
for more information.
|