Kerstin Sophie Haring

The influence of robot appearance on assessment

This paper presents the influence of robot appearance on perception. The goal is to come to an initial understanding of what design and features of robots affect people. We explore the differences how people perceive pet, service, humanoid and android robots. The associations, associated tasks, perception, fears and expectations towards the four different types are evaluated.

Perception of an android robot in Japan and Australia: A cross-cultural comparison

This paper reports the results from two experiments, conducted in Japan and Australia, to examine people’s perception and trust towards an android robot. Experimental results show that, in contrast to popular belief, Australian participants perceived the robot more positive than Japanese participants. This is the first study directly comparing human perception of a physically present android robot in two different countries.

Changes in perception of a small humanoid robot

Humanoid robots are designed to interact with people. To improve the design and development of robots for social human-robot interaction, it is important to consider how people perceive the appearance and behavior of these robots. This paper presents the results of a study on the perception and the changes after passive and active interaction with a physically present humanoid robot in terms of anthropomorphism, animacy, likeability, perceived intelligence and perceived safety. Experimental results show that the perception of the robot changes mainly after the first interaction. The robot is perceived highly likeable in passive interactions with an increase on its perception of animacy.

Perception and Trust Towards a Lifelike Android Robot in Japan

This paper reports the results from an experiment examining people’s perception and trust when interacting with an android robot. Also, they engaged in an economic trust game with the robot. We used the physical distance to the robot, and questionnaires to measure the participants’ character and their perception of the robot. We found influences of the subject’s character onto the amount sent in the trust game and distance changes over the three interaction tasks. The perception of the robot changed after the interaction trials towards less anthropomorph and less intelligent, but safer.

The Use of ACT-R to Develop an Attention Model for Simple Driving Tasks

The use of ACT-R to develop an attention model for simple driving tasks Kerstin Sophie Haring (ksharing@fennel.rcast.u-tokyo.ac.jp) Katsumi Watanabe (kw@fennel.rcast.u-tokyo.ac.jp) Research Center for Advanced Science and Technology, The University of Tokyo 4-6-1, Komaba, Meguro-ku, Tokyo, 153-8904, Japan Marco Ragni (ragni@cognition.uni-freiburg.de) Lars Konieczny (lars@cognition.uni-freiburg.de) Center for Cognitive Science, University of Freiburg Friedrichstr. 50, 79098 Freiburg, Germany of these highly complex tasks. Vice versa, it also can provide an indication for the future development of a cognitive architecture by showing what cannot be realized yet. Abstract Driving a car is obviously a complex task and the construction of an ACT-R model of human attention while performing this task is similarly complex along multiple dimensions and presents a challenge to architecture and modeler. This work is a first attempt to develop an integrated driver model of attention in the cognitive architecture ACT-R. The model is able to keep a traffic lane, identifies traffic signs and crossroads in a sparse, simulated environment. Keywords: Driver behavior model; cognitive architecture; ACT-R; Attention Introduction Fig. 2: Screenshot of the environment interaction with ACT-R. The red circle indicates the current visual focus of attention of the model. For most of us, driving a car is one of our everyday tasks. But even for experienced drivers, just the task itself it is a cognitive challenging task involving a big range of human senses like sight, hearing, touch and acceleration. And this is not yet considering secondary tasks like talking on the phone or visual distractions like city illuminations. Luckily, most driving task are not as challenging as the Traffic Light Tree in Fig. 1, an artificial scenario by the French sculptor Pierre Vivant. The simulation environment for this model was restricted to the components the cognitive architecture can recognize. Nevertheless, basic driving scenarios simulating human visual attention and driver behavior could be implemented. The screenshot form the driving environment, which was separately implemented in Lisp for this work, shows from top-down another (blue) vehicle, the focus of attention (red circle) and the navigation point (N) to keep the vehicle in the center of the road. This model focuses on basic reference points like the horizon, a leading car, the border and the center line of the road, crossroads and traffic signs. For example, the model of a driver in the screenshot in Fig. 2 sets the focus of visual attention on the outer border of the road which enables it to reevaluate the center for the N point. In the next step, it will shift the focus of attention to the front and (hopefully) detect the car in front. If so, possible next steps could be the comparison of the distance to a (here fixed) safety distance or an overtaking procedure. The here presented cognitive model should simulates through ACT-R human attention while driving in a simplified environment and produces the behavior for scenarios with other cars, crossroads and traffic signs. Fig. 1: The (thankfully not on a crossroad) installed traffic light sculpture by Pierre Vivant. The cognitive architecture Current attempts to model human attention while driving a car are realized in a quite more simple environment, yet they are quite an important first step towards the modeling The ACT-R (Anderson, 1993; Anderson 2007) cognitive architecture proposes artificial, computational processes that aim to act like a human cognitive system. Most of its basic