Cohesive Autonomous Navigation System: Image Processing and Data Management

Abstract

The ability of a robotic system to fully and autonomously interact with its environment is key to the future of applications such as commercial package delivery services, elderly robotic assistants, agricultural monitoring systems, natural disaster search and rescue robots, civil construction monitoring systems, robotic satellite servicing, and many more. An architecture that is conducive to Simultaneous Localization and Mapping (SLAM), path planning, and mission planning is a critical element for a system to be robust enough to handle such applications with true autonomy.

In this work, an architecture that lends itself to cohesive operation of all the aforementioned goals is presented. The key components of this architecture are the data management, image processing, SLAM, and path planning. The architecture works through the implementation of a common core database to represent the environment. The database management tools use k-vector search techniques. Image processing techniques are evaluated in a trade study where a graph-based approach is selected. An outline of the SLAM approach and a description of the path planning algorithm employed are briefly discussed. The Cohesive Autonomous Navigation System (CANS) is successfully implemented at slower than real-time speeds and future work is outlined to achieve a real-time system.