Evaluation of trust in robots: A cognitive approach

The study of Human-Robot interaction faces one of the biggest challenges in measuring the trustworthiness of the robots. The enhancement and the augmentation of the human capabilities using the human robot integration are dependent on the reliability and dependability of the robots. These factors become more significant when the participation of the robot is the human robot integration is active and the cohesion between humans and robots is high. In order to measure the trust and other cognitive parameters of the robot, we have designed trust model in this research paper. This paper evaluates the trust of a customized robot while performing a task using three different algorithms. The algorithms used for the path planning task in this paper are simple artificial neural network; reinforcement based artificial neural network and Situation-Operator Model. The trust model proposed in this paper has been simulated using the results obtained while the robot performed its tasks using the three algorithms. The results show that the trust of the robot increases with each learning cycle thereby indicating that the training of the robot enhances the trust parameter of the robot.