ACT-R/E: An Embodied Cognitive Architecture for Human-Robot Interaction

Abstract

We present ACT-R/E (Adaptive Character of Thought-Rational / Embodied), a cognitive architecture for human-robot interaction. Our reason for using ACT-R/E is two-fold. First, ACT-R/E enables researchers to build good embodied models of people to understand how and why people think the way they do. Then, we leverage that knowledge of people by using it to predict what a person will do in different situations; e.g., that a person may forget something and may need to be reminded or that a person cannot see everything the robot sees. We also discuss methods of how to evaluate a cognitive architecture and show numerous, empirically validated examples of ACT-R/E models.

Simulation and HRI Recent Perspectives with the MORSE Simulator

“…When detailed motion or other actions (such as speaking) is required, cognitive models have often been used [19], and recently also in HRI [22]. It is probably safe to say that such models are still far from general, but already quite useful for specific situations.…”

Trustworthiness of Autonomous Systems

“…A research program on trusted AS ought to incorporate mental models informed from cognitive science to better understand and respond to human thoughts and behaviour. An example of such a research program is the recent work programming a robot with ACT-R/E [49,94], an embodied extension of the ACT-R [3] cognitive architecture. The ACT-R/E implementation takes features of human cognition, such as segmenting time into events and narrative explanation to bring meaningfulness and trust to robot-human relationships.…”

“…robots) to engender implicit trust including how they look, sound and feel. Social robots are designed with big responsive eyes and eyebrows [12], as are mobile, dexterous and social robots (MDS) [11,94]. Some designers have shaped robots like baby animals-such as the harp seal robot PARO [1]-and use biomimetic features such as soft skin for tactile trust [50].…”

Coordination mechanism for integrated design of Human-Robot Interaction scenarios

“…There is an infinite number of ways to implement robot control architectures, whether they are biologically inspired by or related to human cognition (e.g., ACT-R/E [3], ICARUS [4], SOAR [5,6]) or engineered to achieve specific requirements (e.g., DIARC [7]). This makes it difficult to compare them [2], since research on robot control architectures is conducted more as feasibility-type studies (e.g., "is it possible to use X to control the robot Y in task Z" [8,9]).…”