Ph.D. Dissertation Proposal - MGLAIR: A Multimodal Cognitive Agent Architecture

Ph.D. Dissertation Proposal Defense:
Tuesday, 17 August 2010, 13:00 in Bell 242

Abstract

I propose to develop and implement a multimodal, grounded, layered, cognitive architecture for embodied agents. The field of cognitive architectures aims to replicate the types of cognitive functionality present in the human mind. The ability to sense and act simultaneously in different modalities is a key feature of humans and other biological cognitive agents. The proposed architecture advances the state of the art through its general treatment of modalities. This treatment includes a conception of a modality as a configurable unitary resource with its own properties that determine its role within an agent’s sensing and acting capabilities, and mechanisms within the architecture to integrate multimodal sensing and acting with reasoning. The architecture will make possible the creation of more capable and more cognitively plausible agents than can be made using existing architectures without these features. This effort will include developing the theoretical foundations of the architecture; an implementation in software of the architecture as a platform for the development of computational cognitive agents; and a new implementation of the SNePS acting system, which will be an integral part of the overall architecture.


Jonathan P. Bona
Department of Computer Science and Engineering
University at Buffalo, The State University of New York