On the Discrimination of Stiffness during Pressing and Pinching of Virtual Springs Giulia Paggetti, Burak Cizmeci, Cem Dillioglugil and Eckehard Steinbach Chair of Media Technology Technische Universit¨ at M¨ unchen Munich Germany Email: burak.cizmeci@tum.de Abstract—In this paper, we study the perception of stiffness while pressing and pinching virtual springs. These exploration events are highly relevant for telerobotics, but their influence on stiffness perception is largely unexplored. Our study contributes to the understanding of the limits of stiffness perception for human-robot interactions. According to our experiments, mean Weber fraction values between 0.134 and 0.166 are obtained for a reference stiffness value of 200 N/m. Our results also show that the differential sensitivity for stiffness is not significantly affected by the actual exploration event. To test the constancy of the extracted Weber fraction values five further reference stiffness values are evaluated suggesting a constant stiffness sensitivity value in the range of stiffness between 135 and 390 N/m. Keywords—Human computer interaction (HCI), Information systems applications, Multimedia streaming, Computer graphics systems and interfaces, Human haptic perception, Psychophysics I. I NTRODUCTION Several studies have been performed to investigate the perception of stiffness and to determine the corresponding We- ber fraction (WF) values. The variability in the experimental methodologies employed makes it, however, hard to connect the different research results and, consequently, to condense the results into a mathematical model of human stiffness perception. For an overview of the WF values observed under different experimental conditions we refer to [1]. Furthermore, stiffness sensitivity is not constant over the whole intensity range of stiffness values [1], [2], [3], and also it is greatly context-dependent [4], [5], [6]. The starting compliance value of an object [1], [2], the property of its surface (rigid vs. deformable surface) [7], the kind of exploratory event applied (pressing vs. pinching, low vs. high amplitude force, tapping vs. pressing) [1], [2], [8], as well as congruent and incongruent multisensory information, and delay in haptic force feedback [9], [10], [11], [12] can strongly influence the perception of stiffness. Additionally, the first studies performed to investigate the perception of stiffness have used human fingers as direct instruments to explore objects (e.g., [13]), whereas more recent studies have been using mainly haptic devices (e.g., [11]), due to their relevance in telerobotic systems. The use of haptic devices introduces a further influence on the studies of human stiffness perception and considerably, different WF values have been determined for different haptic devices (e.g., [14], [11]). For this reason, further experiments need to be performed and relevant conditions in telerobotics also have to be taken into account (e.g., exploration events and multisensory in- formation). The collected experimental results will lead to a deeper understanding of the processes and limitations of human haptic perception and can be used as a basis for the derivation of new and more generally applicable models. In order to be able to extend the collected results to different teler- obotic systems, it is promising to investigate the WF values of human stiffness perception using one of the currently most advanced master haptic devices (e.g., the Force Dimension Sigma 7 [15]). The thresholds found with a high quality haptic device can be then applied conservatively to less accurate devices. In this study, we investigate the discrimination of percep- tual stiffness during haptic interaction in a simulated virtual environment using a high-quality haptic device. In order to make the virtual scene as realistic as possible to a telerobotic system, we use a virtual model of a real teleoperation system in our lab. We study the stiffness discrimination of virtual springs for both pinching and pressing events and determine the corresponding WF values. In telerobotic systems, pressing and pinching events fre- quently occur. The influence of those events on stiffness perception, however, is a research question that needs to be addressed. In [1], compliance sensitivity for both events has been investigated. Users were asked to explore objects with deformable surfaces by both pinching them, between thumb and index finger, and by pressing the fingertip into their surface. The authors found a different stiffness sensitivity for the two cases, with a higher sensitivity for pressing, which is assumed to be caused by different tactile information. How- ever, when probes (e.g., haptic devices) are used to explore deformable objects with rigid surfaces, the stiffness of the objects can be poorly obtained from tactile information [8]. The pressure distribution and the skin deformation, for a given force, are independent of object stiffness [13]. As a result, tactile information is not sufficient to determinate the stiffness of an object and, therefore, kinaesthetic information (the sense of position and motion of limbs along with the associated forces) is necessary. This suggests that different results might be observed in studies involving haptic devices. The differ- ent applied movements might produce different kinaesthetic signals, but information on the specific role of these signals in stiffness discrimination using a probe during pressing and pinching exploration events, however, is lacking. To the best of our knowledge this issue has not been inves- tigated directly. Perception of stiffness has been investigated