Collaboration in  haptic virtual environments

EVA-LOTTA SALLNÄS

Royal Institute of Technology, IPLab

SHUMIN ZHAI

IBM Almaden Research Center, USA

JIN MOEN

Royal Institute of Technology, IPLab

Handing over objects is a common event during collaboration in face-to-face interaction. We investigated how such an event can be supported when the interaction takes place in virtual space. The three-dimensional haptic collaborative interface was designed as a room with two larger shelves, on top of which six cubes were placed, three on each side (Figure 1.).

Figure 1: A sequence showing two subjects performing a hand off in the collaborative virtual environment.

The room also contained two smaller shelves that served as target areas, underneath the two larger shelves. Two cursors, coloured green and blue, corresponded to the tip positions of the two Phantom probes. In both the haptic and the nonhaptic environment it was possible to grasp a cube by placing the cursor on the cube and then pressing the button on the haptic device. Once grasped, the cube could be moved in the environment. When the button was released so was the cube.The haptic user interface was developed so that all surfaces in the environment were touchable and thus provided haptic force feedback. It was also possible to “feel” gravity, the other user’s impact on an object and the collision between a cube and a shelf. The haptic properties were texture, size, weight and stiffness. All other surfaces in the environment were also haptic with a certain friction and stiffness. In the condition without haptic force feedback, the user could neither feel the cubes, walls, floor nor the shelves in the environment. In that case, the Phantom functioned solely as a 3D mouse without force feedback.

In a formal experiment, subjects passed a series of cubic objects to each other and tapped them at target areas. The subjects performance with and without haptic force feedback was evaluated. A within group design was used in this experiment. Each subject was seated in front of a haptic display system in separate rooms. The haptic and the nonhaptic virtual environment were implemented using Reachin Technologies AB’s API on a Windows 2000 PC. The haptic display systems used in this project consisted of two displays from Reachin Technologies AB with two Desktop Phantom force feedback devices from SensAble Technologies, Inc. (Figure 2). This system provides stereo vision through Stereographics CrystalEyes 3 shutter glasses.

                    

Figure 2: Two persons collaborating in the virtual environment using the ReachIn Display system.

Furthermore, we placed our study in the framework of Fitts’ law and hypothesized that object hand off constituted a collaboratively performed Fitts’ law task. Fitts’ law has traditionally been used as a model for performance of individual tasks such as target pointing. Its contribution to user interface design and evaluation include systematic evaluation of different input devices and quantitative comparison of two styles of interfaces such as crossing-based vs. pointing-based interaction defined as the logarithm of the ratio between object distance and object size.  For the same distance, the greater the object is, the faster one can point at it. Although much more complex than simple target tapping, in hand off such a relation may also hold since the larger the object is, the more relative tolerance it allows in the hand off process in terms of the accuracy the giver has to target at the receiver’ hand, and the accuracy the receiver could “grasp” and hold the object.

Given this background, we studied two issues in a collaboratively performed hand off task:

-            If Fitts’ law is applicable as a task performance model for a collaboratively performed hand off task.

-            If haptic force feedback improves performance when passing a virtual object between two persons.

Our results showed that task completion time indeed linearly increased with Fitts’ index of difficulty, both with and without force feedback. The time required for passing objects did not differ significantly between the haptic and nonhaptic condition. However, the error rate was significantly lower with haptic feedback than without.

Papers:

Sallnäs, E-L., & Zhai, S. (2003). Collaboration meets Fitts' law: Passing virtual objects with and without haptic force feedback. In M. Rauterberg, M. Menozzi & J. Wesson (Eds.), Proceedings of INTERACT´03 (pp. 97-104). Amsterdam: IOS Press.

Jin Kjölbergs (Moen) Examensarbetsrapport: http://www.nada.kth.se/utbildning/grukth/exjobb/rapportlistor/2002/Rapporter02/kjolberg_jin.pdf