Error growth in position estimation from noisy relative pose measurements
Robotics and Autonomous Systems
| Hi-index | 0.00 |
This work presents the first step toward an innovative new navigation framework, based on growing a network of reusable paths, to allow a mobile robot to autonomously explore unmapped, GPS-denied, extreme environments. The paradigm (i) results in closer goal acquisition (through reduced localization error), (ii) allows for effective recovery from dead-ends or unproductive routes, (iii) avoids terrain-assessment artifacts due to map merging, and (iv) eliminates the possibility of a robot being unable to find any safe path from the current pose, even when there is one. Extensive simulation and preliminary hardware results are provided.