Design of a socially intelligent task selection software mechanism for a mobile robot
The interaction people have with many current technologies has similar characteristics to the interactions people have with antisocial people. In the best cases the systems know what to do, but often lack the social intelligence to do it in a socially appropriate manner. As a result, these systems frustrate the people using them and are dismissed even though they can be useful. This is a problem given that some of the most exciting new applications for robots require them to cooperate with humans as capable and socially savvy partners. To provide a person with the right kinds of assistance at the right time, a robot partner must not only recognize what the person is doing (i.e., his observable actions) but also understand the intentions or goals being enacted. What was needed was a model for social interaction that was easy to implement and allowed the robot to have knowledge of social interaction with people. Toward these goals, this work attempted to add to existing work in providing awareness to artificial systems, by using the state of interaction as an underlying frame of reference for overall robot performance. The results of this thesis have shown that the implementation of the decision making mechanism was sufficient. This resulted in a social interaction mechanism that provided: (1) Knowledge of Social Context; (2) Socially Appropriate Decision Making Appropriateness. The work presented in this paper was developed in the context of an ultimate goal to develop a type of assistant or personal robot that can be used for high-level tasks.
Alice C Diggs,
"Design of a socially intelligent task selection software mechanism for a mobile robot"
ETD Collection for Tennessee State University.