Cognitive Robots (Physically Situated Agents)

The idea is to develop interesting robotic testbeds that demonstrate the effectiveness of this research in the domain of robots. The first research objective was to establish how the architectures developed in the work on: Agents And Cognition and Computational Architectures map from simulation to physically situated agents.

Of particular interest is the linking of perception, action and thought (motivation, affect and rationality) through the use of heterogeneous, distributed mobile platforms (based on AmigaBots,a Pioneer2, a Pioneer P3DX, Mindstorms, microsensors and a cognitive architecture).

We developed an Omni-Directional (one camera-one mirror) panoramaic vision system for global viewpoints and obstacle avoidance. The data from this sensory apparatus, sonar arrays and other cameras provides fuel to investigate the generation of (egocentric and alocentric) maps, the anchoring problem and cognitive vision.     The   A-CRIBB and motivational construct research   provides a cognitive framework which is being brought together in a subsumption architecture with an extended version of the SRI Saphira architecture [ Konolige et al ].