Simultaneous Localization and Mapping (SLAM)
- Summary
A critical open problem in the development of autonomous mobile robots, such as Mars rovers and unmanned ground vehicles (UGV's) for Earth-based applications, is how to maintain accurate knowledge of the position of the vehicle. We design simultaneous localization and mapping estimation schemes for tracking the position of a robot during long traverses when absolute positioning information is not available. This research focuses on the investigation of the convergence properties of various estimation schemes parameterized on (i) the characteristics of different types of environments (in terms of feature/landmark density and diversity) and (ii) the dead-reckoning capabilities of different types of robots (in terms of the type and quality of the sensor payload).
