The fundamental docking problem in autonomous vehicle and robotic end-effector navigation applications refers to on-line motion path planning to achieve a desired docking pose (position and orientation) within required tolerances. This characterization is evident in numerous approaches used today for the navigation of autonomous vehicles and robotic end-effectors in terrestrial (including manufacturing), air, underwater, and space applications. In this context, high-precision task-space sensors, via passive- or active-sensing techniques, have commonly been used. Frequently, however, a vehicle’s pose cannot be determined accurately due to the inability of task-space sensors to measure orientation as precisely as position. This Thesis addresses this drawback by proposing the utilization of guidance-based methodologies to accurately maneuver the vehicle to its desired docking pose using indirect proximity measurements.

The overall thesis objective is to develop on-line path-planning methods that are not task specific (i.e., generic in nature) and which can be applied to varying vehicle-mobility requirements:​ the use of guidance-based, closed-loop, line-of-sight task-space sensory feedback is proposed for use, in the absence of direct proximity data, for the docking of autonomous vehicles/robotic end-effectors. In this context, two generic guidance methodologies have been developed:​ (i) a model-independent method, which provides effective and accurate guidance that is independent of the sensing-system’s calibration model, (ii) a model-dependent method, which provides guidance by utilizing the sensing-system’s calibration model for accurate docking with a comparably faster rate of convergence. Both methodologies may utilize both passive- or active-sensing schemes to provide corrective vehicle motions. Comprehensive simulation and experimental studies that were carried out demonstrate the advantages that are inherent in utilizing the proposed overall guidance methodologies to address the fundamental docking problem.