Distributed sensor fusion for object position estimation by multi-robot systems

Abstract

Abstract-We present a method for representing, communicating and fusing distributed, noisy and uncertain observations of an object by multiple robots. The approach relies on re-parameterization of the canonical twodimensional Gaussian distribution that corresponds more naturally to the observation space of a robot. The approach enables two or more observers to achieve greater effective sensor coverage of the environment and improved accuracy in object position estimation. We demonstrate empirically that, when using our approach, more observers achieve more accurate estimations of an object's position. The method is tested in three application areas, including object location, object tracking, and ball position estimation for robotic soccer. Quantitative evaluations of the technique in use on mobile robots are provided.

Cooperative situation assessment in a maritime scenario

“…Their on-board sensors and different location in the environment, provide them with different perceptions and executing a complex plan often involves task to be executed by several robots. Hence, in this settings cooperative perception is often used to deal with perception limitation of the single robot [7,43,51].…”

Robot Controllers for Highly Dynamic Environments with Real-time Constraints

“…They report a mean error of 38 cm for a moving ball while the robots did not move. Stroupe et al (Stroupe et al 2001) represent each ball estimate as a two-dimensional Gaussian in a canonical form. This allows to merge the single estimates of the robots simply by multiplying them.…”

Handling Communication Restrictions and Team Formation in Congestion Games

“…Fox has shown that when the amount of information that a robot receives is restricted, teams of robots, with different sensors, can work together to solve the robot localization problem [14]. In addition it has been shown that teams can share sensor information to estimate unobservable parts of the system in robotic soccer domains [30]. Significant work has also been done on coalition formation in distributed systems for the purpose of increasing efficiency, where coalitions are formed dynamically as the system runs.…”