Model-predictive asset guarding by team of autonomous surface vehicles in environment with civilian boats

In this paper, we present a contract-based, decentralized planning approach for a team of autonomous unmanned surface vehicles (USV) to patrol and guard an asset in an environment with hostile boats and civilian traffic. The USVs in the team have to cooperatively deal with the uncertainty about whic...

Full description

Saved in:
Bibliographic Details
Published inAutonomous robots Vol. 38; no. 3; pp. 261 - 282
Main Authors Raboin, Eric, Švec, Petr, Nau, Dana S., Gupta, Satyandra K.
Format Journal Article
LanguageEnglish
Published Boston Springer US 01.03.2015
Springer Nature B.V
Subjects
Algorithms
Allocations
Artificial Intelligence
Autonomous
Boats
Computer Imaging
Computer simulation
Control
Engineering
Mechatronics
Missions
Pattern Recognition and Graphics
Robotics
Robotics and Automation
Robots
Surface vehicles
Tasks
Unmanned vehicles
Utilities
Vision
Decentralized planning
Unmanned vehicles
Unmanned surface vehicles
Task allocation
Online AccessGet full text

Cover

More Information
Summary:In this paper, we present a contract-based, decentralized planning approach for a team of autonomous unmanned surface vehicles (USV) to patrol and guard an asset in an environment with hostile boats and civilian traffic. The USVs in the team have to cooperatively deal with the uncertainty about which boats pose an actual threat and distribute themselves around the asset to optimize their guarding opportunities. The developed planner incorporates a contract-based algorithm for allocating tasks to the USVs through forward simulating the mission and assigning estimated utilities to candidate task allocation plans. The task allocation process uses a form of marginal cost-based contracting that allows decentralized, cooperative task negotiation among neighboring agents. The task allocation plans are realized through a corresponding set of low-level behaviors. In this paper, we demonstrate the planner using two mission scenarios. However, the planner is general enough to be used for a variety of scenarios with mission-specific tasks and behaviors. We provide detailed analysis of simulation results and discuss the impact of communication interruptions, unreliable sensor data, and simulation inaccuracies on the performance of the planner.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0929-5593
1573-7527
DOI:10.1007/s10514-014-9409-9