News Archives

RSS Feed

  • UNM
  • >Home
  • >News
  • >2014
  • >October
  • >[Colloquium] PANTHER: Pattern ANalytics To support High-performance Exploitation and Reasoning

[Colloquium] PANTHER: Pattern ANalytics To support High-performance Exploitation and Reasoning

October 17, 2014

Watch Colloquium:

MOV FILE

  • Date: Thursday, October 16, 2014
  • Time: 11:00 --- 12:15 PM
  • Place: Dane Smith Hall 125


Speakers: Christopher G. Valicka and John M. Linebarger, Sandia National Laboratories

It is well known that the amount of sensor data being generated and monitored today is enormous, and analyzing the data efficiently and accurately in order to extract the pertinent information is a significant challenge. Sandia National Laboratories' PANTHER (Pattern ANalytics To support High performance Exploitation and Reasoning) project is working to help solve this problem by turning overwhelming geospatial data streams into meaningful connections and patterns.

Because current approaches focus mainly on pixel based analysis, they lack the ability to establish relationships between the differing features and events that are generated from the various sensor types, specifically those based on either space or time.There is a need to not only associate spatial and temporal data also but to enable the recognition and anticipation of relationships within incomplete data and to assign meaning to these relationships such that the most complete picture can be synthesized and understood.

PANTHER aims to establish a more sophisticated standard, shifting from the current pixel based analysis of the image to pattern analysis of primitives within the image, in order to isolate temporal and spatial behaviors.In working to alleviate the present inadequacies, researchers are developing new computational geometry and machine learning algorithms for the analysis of motion/trajectory behaviors, testing these algorithms on problems drawn from national security issues, and evaluating them with respect to both computational and human oriented metrics.

In this presentation we intend to describe the problem space, summarize existing work, characterize the research contribution and apply it to several examples in the domain, and suggest future work.

About the speakers:
*Christopher G. Valicka, PhD.* Chris received his Ph.D. in Systems Engineering from The University of Illinois at Urbana-Champaign. His thesis work focused on multi-objective control and coordination of multi-robot systems. Chris's interests include numerical optimization, multi-agent dynamical systems and software development for the control of autonomous systems. He currently works in Tasking, Planning, and Mission Management as part of the Systems Mission Engineering Center.

*John M. Linebarger, PhD, MBA*.John has been with Sandia for 25 years in the areas of Information Systems and Computer Science.His PhD is from Lehigh University, and his areas of expertise include distributed systems, software engineering, scientific visualization, and project management. Immediately after the colloquium Sandia will be hosting a Meet 'n Greet pizza lunch in the main conference room on the first floor of the ECE building (ECE 118), from 12:30 until 2:30. Free pizza and soda; opportunities to speak with Sandia hiring managers; and a chance to see a brief presentation about Sandia National Labs and the student intern hiring pipeline we would like to establish with UNM.