Boer, B. de and Rijn, H. van and Wijbenga, A. (2007) AIBO-R: integrating cognitive models and robotics. Doctoral, Artificial Intelligence.
|
Text
AIBO-R.pdf - Published Version Download (941kB) | Preview |
Abstract
Over the years, the field of robotics has primarily focused on low-level behaviour, like navigation and exploring, while cognitive modelling has primarily focused on high-level cognition. As a result, each field often disregards the other field’s level of cognition. Combining a cognitive model with a robot, such that multiple levels of cognition are implemented, might be beneficial for both robotics and cognitive modelling. To explore the possible benefits of that combination, an embodied cognitive model of route learning was developed. The model was developed using ACT-R, which has been expanded to interface with a Sony AIBO robot. The study shows that creating an interface between a cognitive model and a robot is challenging, especially when low- and high-level cognition have to interact in a plausible and useful way. From the cognitive modelling perspective, it forces one to include several aspects which in conventional models might have been overlooked. At the same time, the study showed that robotics could benefit from the unified representations and learning mechanisms of a cognitive architecture, which result in useful top-down control. As a result, a robot is able to learn in a cognitively plausible way and is able to improve its performance by adapting to a new environment.
| Item Type: | Thesis (Doctoral) |
|---|---|
| Degree programme: | Artificial Intelligence |
| Thesis type: | Doctoral |
| Language: | English |
| Date Deposited: | 15 Feb 2018 07:28 |
| Last Modified: | 15 Feb 2018 07:28 |
| URI: | https://fse.studenttheses.ub.rug.nl/id/eprint/8459 |
Actions (login required)
 |
View Item |