Intermittent Measurement in Robotic Localization and Mapping with FIM Statistical Bounds

Abstract

The focus of study is to review the FIM statistical behavior in each EKF update and determine its potential in providing sufficient information about Robotic Localization and Mapping problem with intermittent measurements. We provide theoretical analysis and prove that the FIM can successfully describe both upper and lower bounds for the state covariance matrix whenever measurement data is not arrived during robot observations. This approach can give a better picture on how information are processed in EKF when measurement data is partially unavailable. Some simulation evaluations are also included to verify our results and consistently demonstrate the expected outcome.