Building and sharing the cognition model of the environment with collaborative service robots
Loading...
URL
Journal Title
Journal ISSN
Volume Title
Sähkötekniikan korkeakoulu |
Doctoral thesis (monograph)
Checking the digitized thesis and permission for publishing
Instructions for the author
Instructions for the author
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Authors
Date
2011
Major/Subject
Mcode
Degree programme
Language
en
Pages
Verkkokirja (18457 KB, 123 s.)
Series
Aalto University publication series DOCTORAL DISSERTATIONS ,
104/2011
Abstract
This research presents a shared cognition concept for building a cognition model collaboratively by robot and human. The proposed concept is based on an object-oriented approach that abstracts robot's perception, cognition, and knowledge to separate but interconnected elements. The shared cognition concept allows the human and robot share information related to objects in the environment for effective task execution and information exchange. The concept enables building a model using sensors, as well as inputting data from the user. User can utilize also robot's sensor data for inputting information. This utilizes the robot's perception of the environment in conjunction with human's cognitive understanding. The proposed concept can be used for exchanging information on the objects, their classes, identities, appearances, locations, structures, and states. Representation of perceived entities is done through observed objects, robot's cognitive understanding of the objects is represented by real objects, and the knowledge of objects' classes by meta-objects. Observed objects are created with segmentation which can be autonomous or assisted by human. Real objects are used to transfer the actual object information between human and robot. Recognition of objects is done by matching real and observed objects. Uncertainties of recognition and localization are handled by probabilistic approach. The model enables learning from perceived information allowing the model to improve its performance as more data is gathered. The cognition model can be used for collaborative task execution. Several aspects of the model are validated with two different platforms, WorkPartner service robot and Avant machine based semi-autonomous robot. Based on the tests, the proposed model is an effective mechanism for collaboratively exchanging spatial information between a robot and a human.Description
Supervising professor
Halme, Aarne, Prof.Thesis advisor
Halme, Aarne, Prof.Keywords
service robotics, shared cognition, robot cognition, robot perception