JEKATERINA NOVIKOVA

Your ePortfolio

Project Background

Project Background

The problem, analysed in the thesis, belongs to the cognitive engineering area, the area which brings together cognitive science (including artificial intelligence), engineering methodologies and practices, information science and design, information retrieval and similar approaches. The multi-disciplinary focusing on cognitive models and real-time embedded systems, such as mobile robots, helps to emerge a broader and deeper understanding of robotics as part of everyday life and society.

An integrated cognitive architecture can be defined as a single system that is capable of producing all aspects of behaviour, while remaining constant across various domains and knowledge bases, thus modelling human performance in multimodal multiple task situations (Newell 1990; Anderson et al. 2004). Allen Newell in his 1990 book Unified Theories of Cognition provided 12 criteria for evaluation of cognitive systems: adaptive behaviour, dynamic behaviour, flexible behaviour, development, evolution, learning, knowledge integration, vast knowledge base, natural language, real-time performance, and brain realization. These criteria have been analyzed and applied to ACT-R, Soar and classical connectionist architectures (Duch, W. et al., 2008).

Research on integrated cognitive architectures is interdisciplinary by nature, spanning the fields of artificial intelligence, cognitive psychology, and neurobiology. Over the past decades, many cognitive architectures have been proposed and steadily developed, based on different approaches and methodologies, such as Soar, ACT-R, ICARUS, BDI, CLARION and others.

Although the impressive progress in many specific sub-topics in AI and Cognitive Science is recognized, work on building integrated cognitive systems moves slowly. All the current cognitive architectures are far from the goal to cover the requirements for general intelligence. Most systems able to perform complex tasks that humans and other animals can perform easily, have to be very carefully crafted, normally their field of expertise is very narrow, and they are hard to extend. Whatever intelligence they have could be described as 'insect-like', with very little flexibility or self-understanding. One of the possible reason for this is that over the last years research has become highly specialized: many individuals and research teams focuse their efforts on narrow specific problems, such as vision, or learning, or language processing, or problem solving, or mobile robotics.

The main purpose of this master project is to make one step further towards developing a cognitive system that is able to work fast and efectively in open ended, challenging environments, dealing with novelty, uncertainty and change.

 
Powered by: itslearning