Multi-UAV target search with expected observation

Multi-UAV target search with expected observation

Lanillos, P., Gan, S. K., Besada-Portas, E., Pajares, G., Sukkarieh, S., (2014). Multi-UAV Target Search using Decentralized Gradient-based Negotiation with Expected Observation. Information Sciences, Volume 282, 20 October 2014, Pages 92-110, ISSN 0020-0255

JCR, impact factor 2.833, position 9/125, Q1, category Computer Science and Information Systems.

matlab source code
in Matlab of an example of the Lost Target Search using Expected Observation (simplified and centralized version of piecewise linear optimization for searching a static lost target)


A novel information-based cooperative and decentralized algorithm for searching lost targets that transforms a local decision making into an informed one reducing the searching time. It uses an expected observation heuristic to estimate the information that the agents will get. The novelty lies in modeling the heuristic as a sensor that makes easy the construction of the utility function and yields to a fast execution of the algorithm. This work will improve many real applications that involves information uncertainty such as: Search and Rescue Missions for airplane and ship-wreck accidents, long-term learning in unknown environments, etc; it will help the Unmanned Air Vehicles and the Unmanned Ground Vehicles community to have real time and close-optimal algorithms; and it will provides a new way to improve any Information-based decision making problem.

Bibtext source:

title = “Multi-UAV target search using decentralized gradient-based negotiation with expected observation “,
journal = “Information Sciences “,
volume = “282″,
number = “0″,
pages = “92 – 110″,
year = “2014″,
issn = “0020-0255″,
doi = “″,
url = “″,
author = “Pablo Lanillos and Seng Keat Gan and Eva Besada-Portas and Gonzalo Pajares and Salah Sukkarieh”,
keywords = “Cooperative search”,
keywords = “Decentralized decision making”,
keywords = “Probabilistic reasoning”,
keywords = “Unmaned air vehicle ”