Multirobot cooperative event based localization scheme

This video presents the experimental results obtained with a new method of multirobot cooperative event based localization scheme that improves the bandwidth usage in a heterogeneous group of mobile robots. This method relies on an agent based framework that defines the communications between robots and in an event based Extended Kalman Filter that performs the sensor fusion from local, global and relative sources. The event is generated when the pose error covariance exceeds a predefined limit (RA,lim).  By this, the robots update the pose using the best relative information available only when necessary, using less bandwidth and computational resources when compared to the traditional time based method, allowing bandwidth allocation for other relevant tasks while extending battery life.

The first test shows a group of five robots following a linear trajectory, and compares the time based approach (TCLA) with the event based approach (ECLA) , both being error bounded solutions, but in the case of the ECLA, with lower bandwidth usage.

The second test shows a group of ten robots advancing at constant speed while employing a Braitenberg algorithm to avoid any obstacles  (neighbor or a wall). Four cases of event limit (RA,lim) are compared  to show its influence in the localization algorithm behavior.  The final figures show a stable behavior and bounded error with all the used RA,lim values. Also the tradeoff between performance and the communications is observed as the accuracy is improved with lower limit values and decreasing the transmitted messages with higher limit values.