[up]

e-Merge-ANT Presentations

AAAI 2002, Edmonton, Alberta, Canada

Main presentation: PDF or HTML
Presentation of a paper on soft graph coloring of dense graphs at a workshop on Probabilistic Approaches in Search at Eighteenth National Conference on Artificial Intelligence

AAMAS 2002, Bologna, Italy

Main presentation: PDF or HTML
Presentation of a paper on asynchronous execution of a soft graph coloring algorithm at The Third International Workshop on Distributed Constraint Reasoning at First International Joint Conference on Autonomous Agents & Multiagent Systems

IDPT 2002, Pasadena, California, USA

Main presentation: PDF or HTML
Presentation of a paper on peer-to-peer constraint optimization and sensor coordination at The Sixth Biennial World Conference on Integrated Design & Process Technology

Summer 2002, Captiva Island PI Meeting

Main presentation: PDF or HTML
The main topic of this presentation is Kestrel's approach to optimizing collaborative scan scheduling under uncertain knowledge about target behavior.

Winter 2001, Scottsdale PI Meeting

Main presentation: PDF or HTML
The main topic of this presentation is Kestrel's approach to the second-year challenge problem: optimization of measurement quality through distributed, anytime scheduling combined with distributed tracking.
Negotiation & Challenge Problem meet Complexity & Dynamics: PowerPoint or plain text format
Summary of working group.

SAGA 2001, Berlin, Germany

Main presentation: PDF or HTML
Presentation of a paper on soft graph coloring at 1st Symposium on Stochastic Algorithms, Foundations and Applications

Spring 2001, Tahoe City PI Meeting

Main presentation: PDF or HTML
The main topic of this presentation is decentralized, approximate graph coloring applied to distributed resource management.
Animations (AVI files)
These illustrate the algorithms discussed in the main presentation.

Autumn 2000, Charleston PI Meeting

Main presentation: PDF or HTML
This presentation deals with a decentralized algorithm for resource management that is based on graph coloring, and presents experimental results based on the ANTs challenge problem (distributed radar tracking).
Animation AVI
This shows the distributed tracking at work.
  • There are 4 radar units, one at each corner - when the radar heads are active, they are displayed brightly.
  • The blue sphere shows the true path of the target - it moves along a diamond path, the corners of which are marked (in blue).
  • The little colored squares that fade in and out show the tracker's estimates of the target's position. They are green if the estimate is accurate, red if it is inaccurate and yellow/orange inbetween.
  • The sphere with the attached arrow represents an estimate of the target's position computed by interpolating between the discrete track esimates (see previous item). The arrow represents an estimate of the target's velocity (direction and magnitude). The color coding is the same as the previous item (green=good, red=bad).

Spring 2000, Seattle PI Meeting

Main presentation: PDF or HTML
This presentation lays out the framework for our approach to distributed resource management through scheduling.