Towards a Cloud-Based Architecture for 3D Object Comprehension in Cognitive Robotics

Abstract

Abstract—Cognitive robotics can take advantage of distributed, web-based information as a foundation for comprehending 3D objects in a 3D scanned world. The proposed CogOnto model makes possible grounding a cognitive computing system with sensor data gathered from diverse and heterogeneous sources, associated with humanly crafted symbolic descriptors. The system supports cognitive embodiment within the totality of an information ecology, and not just within the physical world where an individual robot, essentially a mobile peripheral device, is located. The informed system uses 3D objects as common denominators for shared world comprehension.