Social Robot Intelligence Quotient to Measure Human-Robot Interaction Performance for Human-Care Service Robots

Abstract

In order to measure human-robot interaction (HRI) performance for human-care service robots, a concept of social robot intelligence quotient (SRIQ) is proposed, which is composed of five element intelligence quotients (EIQs): Human Recognition, User Modeling, Body Movement, Self-Expression, and Communication. The SRIQ is assessed as the weighted sum of the EIQs where the weight factors are determined according to the morphology of the robot subject. Each EIQ is also assessed as the weighted sum of the scores for its component skill factors which actually measure their performance. Preliminary evaluation tests are conducted for the SRIQ elements with a robot subject called Jenny and the test results verify the feasibility of the SRIQ as a system-level evaluation measure for the social robot intelligence.