 |
Exploits all-source
data to identify, characterize, and localize threats
without highly trained operators or prior threat characterization |
|
Operates on multiple
and disparate data sources to identify suspicious events
embedded within complex backgrounds |
|
Trains on tracking
and surveillance data to identify typical behaviors in
the surveillance region |
|
Applies innovative
unsupervised machine learning techniques to understand
the environment and detect anomalous occurrences within
vast data sets |
Applications |
| DISCOVER-ie
has broad applications to federal, state, and local agencies
involved with public health and homeland security as
well as Department of Defense situational awareness initiatives.
It has been successfully demonstrated in disparate applications,
including: |
|
Port and shoreline
security |
|
Covert shoreline operations
detection |
|
Maritime domain awareness |
|
Bioterrorism surveillance |
|
Medical database monitoring |
|
Highway monitoring |
|
Telecommunications monitoring |
|
Sensor fusion |
Features |
|
Flexible and automated
surveillance system |
|
Handles large, diverse set of
real-time data |
|
Anomalous event diagnostics to
allow adjudication of alerts |
|
Allows, but does not require,
prior knowledge of threat behaviors |
|
Maintains sensitivity to weaker
anomalies by adapting to changing background levels in
ambient patterns/behaviors |
|
Provides operator control over
alert rate |
|
Supports post analysis |
Benefits |
|
Reduces costs while
providing full data exploitation |
|
Reduces operator training and
manpower requirements through automatic identification
of potential threats |
|
Turns multi-source data into knowledge |
