/AU_ROS_UAV

A collision avoidance system for the Auburn University EasyStar UAV project

Primary LanguageC++

  This is a collision avoidance module for the Auburn University UAV ATTRACT
  project. It uses Dynamic Sparse A* Search to plan optimal or nearly-optimal, 
  flyable paths for UAVs.
  
  This code is a collaborative effort between Thomas Crescenzi of Marist College,
  Tyler Young, and Andrew Kaizer, both of Truman State University.

  It was created during an 8-week, NSF-funded Research Experience for 
  Undergraduates (REU) at Auburn University, between May and July of 2011.
  
  As it stands, the project sadly does not live up to its full potential. Whereas  
  we confirm that A* can safely and reliably guide up to 15 aircraft in a 500 m 
  square airspace in simulations without tight turning constraints, enforcing real-
  world turning capabilities has proven much more time consuming than we anticipated.
  Given more time, we are confident these results could be obtained in realistic 
  simulations. As it stands, however, our simulations show 1 to 2 potential 
  collisions for as few as 8 aircraft in a 500 m by 500 m airspace, with as many 
  or more for 16 aircraft in the same space.