Advertisement

Abstract

The book is concluded by a chapter summarizing the major findings and showing the possible extensions to the research presented in the previous chapters.

10.1 Summary

This book is the first attempt to formalize the specific artifacts corrupting the rendering of virtual haptic textures. At a first glance, this document can be read as a practical guide for precise haptic textures; this is partially the intent of the author: to offer a set of simple conditions to guide haptic researchers towards artifact-free textures. The conditions identified in this work are also extremely valuable when designing psychophysical experiments (because not all the textures can be rendered on a haptic device) and when analyzing the significance of the data collected.

10.2 Results

The guidelines of this book, however, are clearly motivated and, for the most part, experimentally validated; moreover, the passivity conditions and the characteristic number, required a novel interpretation of the same idea of passivity when applied to multidimensional virtual environments.

10.2.1 Passivity

The characteristic number is a measure of the impedance of virtual haptic texture algorithms, and, coupled with the novel interpretation of passivity, prevents control related artifacts. Alternative approaches are available in the literature, among the others, virtual coupling could stabilize the interaction, both for conservative and non-conservative force fields [1]. Another viable solution is the passivity observer with energy following, which, in real time, ensures the energy balance of the haptic interaction [3].

These on-line approaches, however, do not offer any insight on the nature of the unstable texture, hence their application could remove perceptually significant aspects of the textured force field. On the other hand, the characteristic number matches the maximum impedance renderable by the haptic device with the impedance of the virtual textures, which leads to more deterministic force fields, better suited for psychophysical studies.

10.2.2 Devices and Algorithms

The combination of the Pantograph and the new friction based algorithm is a well characterized experimental setup for the study of the perception of textures.

The oversample and filter approach is pivotal for the quality of the haptic textures rendered with the Pantograph, but alternative solutions exist. The acceleration matching technique, successfully applied to thePhantom , imposes precise open loop acceleration profiles to the handle of the device, once the dynamic model of the combination device/user is identified. At this stage of development, acceleration matching seems to be better suited for rendering time-varying stochastic textures, because its applicability to closed-loop multidimensional virtual environments has not yet been investigated [4].

The most evident limitation of the Pantograph is the 2D workspace and the 2D forces it generates. Among the multidimensional devices, however, no device offers the same level of resolution, bandwidth, low inertia, and low friction. The only possible exception is the Ministick, whose frequency response, though, has not yet been measured [2].

The proposed friction based algorithm is physically inspired and plausible, and easily extendable to 3D curved surfaces. In addition, it has a number of parameters to fine tune the texture sensation and the passivity of the interaction, and its dissipative nature is instrumental for the rendering of a stable texture. Friction maps recorded from the surface of real objects, e.g., [5], can be rendered with this algorithm without modifications; moreover, an analysis of the gradient of those maps would ensure passive rendering. Finally, the extension to 3D objects of friction based approaches has a clear advantage over the geometry based methods, because the collision detection and the computation of the minimal distance are performed with the low curvature surface and not with the texture boundaries. This simplification is important, for example, when the applied texture is a profile measured from a real surface, because the collision detection algorithm is not affected by the potential complexity of the profile.

The innovations of this work are clearly targeted to the most demanding audience in the haptic community, those researchers who want to generate repeatable stimuli for the study of psychophysics of touch. The same techniques, however would benefit also the less demanding haptic applications: the parameters of the novel texture algorithm can be tweaked to account for the specification of most haptic devices; moreover, the characteristic number does apply to any texture algorithm regardless of the quality of the haptic device.

10.3 Future Work

The most enticing application of the work presented in this book is the psychophysics study of indirect touch. The innovations presented here finally permit to generate textural stimuli with guaranteed quality, ideal for the study of perceptual properties of virtual textures, among which roughness is of great importance. For example, it is possible to apply the conditions identified here to geometry profiles sampled from real surfaces, for the study of the perception of real versus virtual textures.

A second possible avenue of extension of this work, regards the application of haptic textures to virtual reality. The novel formulation of the friction based algorithm and, its three dimensional extension, would be the ideal candidate for adding textures to generic curved surfaces, without affecting the underlying computational engine. For example, in a surgical simulator, the texture of a virtual bone could be changed to express different states of decay of the tissues, without the need of a complex underlying geometry. On the other hand, the surfaces are in general represented with triangular meshes whose discontinuities are known to generate perceptual artifacts. These discontinuities are usually handled by “blending” the normals of the triangles, thus creating a curvature on the surface. The extension of the analysis in Chap.  7 to non-regular surfaces, such a blended triangular meshes, is left for future work.

References

  1. 1.
    Adams, R.J., Hannaford, B.: Stable haptic interaction with virtual environments. IEEE Trans. Robot. Autom.15(3), 465–474 (1999) CrossRefGoogle Scholar
  2. 2.
    Cholewiak, S., Tan, H.Z.: Frequency analysis of the detectability of virtual haptic gratings. In: WHC ’07: Proceedings of the Second Joint EuroHaptics Conference and Symposium on Haptic Interfaces for Virtual Environment and Teleoperator Systems, pp. 27–32. IEEE Computer Society, Washington (2007) CrossRefGoogle Scholar
  3. 3.
    Hannaford, B., Ryu, J.H.: Time-domain passivity control of haptic interfaces. IEEE Trans. Robot. Autom.18(1), 1–10 (2002) CrossRefGoogle Scholar
  4. 4.
    Kuchenbecker, K.J., Niemeyer, G.: Improving telerobotic touch via high-frequency acceleration matching. In: Proceedings of the IEEE Int. Conf. on Robotics and Automation (2006) Google Scholar
  5. 5.
    Pai, D.K., van den Doel, K., James, D.L., Lang, J., Lloyd, J.E., Richmond, J.L., Yau, S.H.: Scanning physical interaction behavior of 3D objects. In: SIGGRAPH ’01: Proceedings of the 28th Annual Conference on Computer Graphics and Interactive Techniques, pp. 87–96. ACM Press, New York (2001) CrossRefGoogle Scholar

Copyright information

© Springer-Verlag London Limited 2011

Authors and Affiliations

  1. 1.MontrealCanada

Personalised recommendations