Intrinsically Motivated Autonomy in Human-Robot Interaction: Human Perception of Predictive Information in Robots

Abstract

In this paper we present a fully autonomous and intrinsically motivated robot usable for HRI experiments. We argue that an intrinsically motivated approach based on the Predictive Information formalism, like the one presented here, could provide us with a pathway towards autonomous robot behaviour generation, that is capable of producing behaviour interesting enough for sustaining the interaction with humans and without the need for a human operator in the loop. We present a possible reactive baseline behaviour for comparison for future research. Participants perceive the baseline and the adaptive, intrinsically motivated behaviour differently. In our exploratory study we see evidence that participants perceive an intrinsically motivated robot as less intelligent than the reactive baseline behaviour. We argue that is mostly due to the high adaptation rate chosen and the design of the environment. However, we also see that the adaptive robot is perceived as more warm, a factor which carries more weight in interpersonal interaction than competence.

Warmth and Competence to Predict Human Preference of Robot Behavior in Physical Human-Robot Interaction

“…Figure 1 shows the robot and the experimental environment of the study. We use the spherical robot Sphero in its BB8 version [14], [15]. The head is attached to the robot's inner vehicle with a magnet, providing the participant with a sense of the robot's direction.…”

“…b) Updating Network Weights: In condition A and C, the robot was equipped with a computational model of intrinsic motivation [17]. The update rules for the network are implemented by time-local predictive information [14], [18], [19]. The robot tries to excite different sensors through the generation of a variety of motion regimes, but in a predictable way.…”

Human Perception of Intrinsically Motivated Autonomy in Human-Robot Interaction

“…This section describes the robot and the used evaluation measures, both used in a preliminary study (Scheunemann et al, 2019, described in the next section) and the follow-up study, the main study of this article, described afterwards. It is argued that an anthropomorphic robot platform may raise expectations of its social capabilities in participants (Dautenhahn, 2004;Hayashi et al, 2010), which in turn may interfere with the investigation of the perception of intrinsic motivation.…”

“…The next sections are brief descriptions of the study design and its results and implications for the current study. More details on the preliminary study are discussed in (Scheunemann et al, 2019).…”

“…Instead, the robot and the human participants explore their behavior towards each other. The preliminary study has been conducted and published prior to this work (Scheunemann et al, 2019). We will summarize the preliminary study, to introduce the first steps towards investigating human-perception of an intrinsically motivated robot, and to present design decisions, like for example the baseline behavior, which is used in both studies.…”