Preliminary Evaluation of a Wearable Camera-based Collision Warning Device for Blind Individuals

This work describes a preliminary evaluation of a wearable collision warning device for blind individuals. The device was found to provide mobility benefit in subjects without (or deprived of) vision. This preliminary evaluation will facilitate further testing of this developmental stage device in more naturalistic conditions.

We developed a wearable video camera-based device that provided tridirectional collision warnings (right, center, and left) via differential feedback of two vibrotactile wristbands. We evaluated its mobility benefit in blind and normally sighted (NS) blindfolded individuals in indoor mobility courses.

Three evaluation experiments were conducted. First, the ability of the device to provide warnings for hanging objects not detected by a long cane was evaluated in eight NS and four blind subjects in an obstacle course with and without the device. Second, the accuracy of collision warning direction assignment was evaluated in 10 NS subjects as they walked toward a hanging object at random offsets and verbally reported the obstacle offset position with respect to their walking path based on the wristbands' vibrotactile feedback. Third, the mobility benefit of collision warning direction information was evaluated by 10 NS and 4 blind subjects when walking with and without differential wristband feedback.

In experiment 1, collisions reduced significantly from a median of 11.5 without to 4 with the device (P < .001). Percent preferred walking speed reduced only slightly from 41% without to 36% with the device (P = .04). In experiment 2, the most likely reported relative obstacle positions were consistent with the actual positions. In experiment 3, subjects made more correct navigational decisions with than without the collision warning direction information (91% vs. 69%, P < .001).

Substantial mobility benefit of the device was seen in detection of aboveground collision threats missed by a long cane and in enabling better navigational decision making based on the tridirectional collision warning information.