An Uncertainty-driven Analysis for Delayed Mapping SLAM

This paper deals with the analysis of the delayed SLAM problem from the perspective of the uncertainties involved in the process. We consider an autonomous mobile robot moving in an environment and equipped with noisy encoders, used for the ego-motion reconstruction, and with a LIDAR for indoor features detection. We adopt an Extended Kalman Filter (EKF) based solution and we analyse the effect of the length of the delayed measurement window on the system uncertainties. The analysis covers the standard LIDAR measurements, but it is also extended to range-only measurements. Mont Carlo simulation results are provided on synthetic indoor environments for both the cases.