Haptics in Bimanual Manipualtion

In virtual reality based rehabilitation, bimanual exercises can be performed by exploiting haptic rendering techniques that allow object manipulation with two haptic devices. Haptic interaction generally involves spherical end-effectors with invariant shapes. During object manipulation in VR, this impacts the ability to determine the real position of the end effector after the collision between it and the haptic interface point.

 Humans synergically exploit both visual and haptic information when performing manipulation tasks. The mechanisms behind visuo-haptic integration have been a topic of great debate for more than a century. Early studies on sensory influence stated that vision was the dominant sense. However, it is now commonly accepted that the two sensory channels are weighted based on their reliability and appropriateness. When spatial interaction tasks are concerned, the visuo-haptic coupling is characterized by strong visual dominance; the sense of touch is favoured in tasks that require the perception of texture and material properties. Accordingly, it is argued that visual cues afford more efficient bimanual handling in VR. 

The present study sought to examine whether additional visual information regarding the penetration of the wrists into the virtual object (i.e., the color and shape of the spheres changed according to the level of force exerted by the subject) leads to improved bimanual task performance in a virtual environment. To this end, six neurologically healthy participants performed an object manipulation task with haptic feedback (haptic condition) and with haptic feedback as well as additional visual cues (haptic+visual condition).

Figure Caption. (a) Wrist robots used in the study; (b) Mapping between virtual environment (VR) and wrist movements. The virtual environment consisted of two spheres, a cardboard carton representing the object to be manipulated (with a superimposed countdown timer), and a semitransparent carton representing the to-be-reached target. The horizontal displacement of the sphere is controlled by wrist flexion/extension (FE), while the vertical displacement is controlled by wrist adduction/abduction  AA); (c) Target positions shown during the experiment. Each position was randomly presented, for a total of 10 repetitions per condition; (d,e) Visual feedback used for the two conditions: (d) when haptic cues were provided the shape of the spheres did not change if a collision occurred, (e)while in the haptic+visual condition the shape and colour of the spheres deformed according to the haptic interface point (HIP) into the virtual object

REFERENCES

 

Contu S, Hughes C and Masia L. "The role of visual and haptic feedback during dynamically coupled bimanual manipulation" IEEE Transactions on Haptics , vol.PP, no.99, pp.1-1.

Sara Contu, Leonardo Cappello, Juergen Konczak and Lorenzo Masia. "Preliminary analysis of non-dominant proprioceptive acuity and interlimb asymmetry in the human wrist"  37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2015) Milan, Italy, August 25-29 2015.

 Sara Contu, Charmayne Hughes, Lorenzo Masia. "Influence of Visual Information on Bimanual Haptic Manipulation". IEEE International Conference on Robotic Rehabilitation 2015, 11-14 August 2015, Singapore