Thursday, April 16, 2009

teaching and reinforcing appropriate cane arc width

teaching appropriate cane arc width
In Rob’s (and Ashmead) 2002 article we find these statements:

“For this study, the widest part of the participants’ bodies were their shoulders. Across groups, conditions, and sessions, arc widths were 38.9 cm (or about 1 ft) wider than the participants’ shoulder widths”

“The participants increased their arc width-to-shoulder width differential when they were first reminded of the proper two-point touch technique, increased it further when they relaxed their technique, and then reduced it when they were reminded of the proper technique the second time. Given the differences in hand position and wrist bending already noted, it appears that the participants were not able to use biomechanical feedback to structure their technique. They erred on the side of caution and used arc widths twice as wide as prescribed so that any other deviation in technique would not shortchange body coverage.

The involvement of wrist movement and hand position is not sufficient to explain the pattern of arc widths across conditions. The movement of horizontal hand position to the right in the Normal and Comfort conditions, with an accompanying lessening of wrist extension, would lead one to expect smaller arc widths in the Normal and Comfort conditions and larger arc widths in the Classic1 and Classic2 conditions. However, arc width-to-shoulder differentials increased from the Normal to the Classic1 to the Comfort conditions and decreased in the Classic2 condition. This pattern of results is different from any of the other measures. It may indicate that arc width is not solely a biomechanical process but is strongly influenced by cognitive demand, more so than other aspects of the long-cane technique."

Does this mean that WiiCane feedback will be an effective tool because it will reinforce via a audio-tatctile-cognitive channel?

We do not have any hard data on teaching the production of an effective cane arc width. Motor-skills learning is well documented, but we have no studies on what constitutes a strategy or approach to reinforcing a target behavior. There is also scant annecdotal data.

Sadowski, J. (2004). Springboard. Re:View, 35(4), 160-164.

She’s Called on the Carpet to Teach Arc Width

I have used several methods to teach arc width. Most give an indication of the arc width by providing a stopping place for the cane on each side of the swing. Two canes extended on the floor, parallel to the line of travel and to each other, or a guide made of PVC pipe, effectively provide information concerning arc width, but they do not teach control of the arc width. I find many students enjoy hitting the boundary objects for the sound feedback and swing harder than is necessary. The result is that when the guide is taken away, the student is not in control of the swing and the arc width is too wide. As a result students may not develop the touch and control needed to demonstrate consistent, correct arc width.

One day while standing in the hall, pondering the problem, I spotted the rugs that were placed outside the classrooms for students to place their winter boots on to dry. The width was about the correct width of a particular student’s arc. We experimented with his standing centered on the carpet and swinging the cane back and forth. The rug edge provided feedback to help him know when to begin the counter arc. It helped him to learn the correct width and to develop the control of the cane that solid obstacle barriersdid not.Longer lengths of carpet can be used to permit the student to walk a distance and practice arcing the cane while moving. This provides an added bonus of feedback concerning straight line of travel.Pieces of rug can be found at carpet store remnant bins. Most dealers are willing to donate a section of carpet when they understand the use. Longer lengths can be used to permit the student to walk a distance and practice arcing the cane while moving. Storage and transport are relatively easy because the carpet pieces can be rolled up and secured with a bungee cord.

Jane Sadowski, COMS
Special Education Districts of McHenry/Lake Counties
1200 Claussen Drive, Woodstock, IL 60098
jafsadowski@ameritech.net

Veering re-visited: Noise and posture cues in walking without sight

I thought that the information about auditory stim and veering was interesting and something to keep in mind when we do a final design on the spoken/sounds feedback system.

Millar, S. (1999). Veering re-visited: Noise and posture cues in walking without sight. Perception, 28(6), 765-780.


Millar S.
Department of Experimental Psychology, University of Oxford, UK. susanna.millar@psy.ox.ac.uk
Effects of sound and posture cues on veering from the straight-ahead were tested with young blind children in an unfamiliar space that lacked orienting cues. In a pre-test with a previously heard target sound, all subjects walked straight to the target. A recording device, which sampled the locomotor trajectories automatically, showed that, without prior cues from target locations, subjects tended to veer more to the side from which they heard a brief, irrelevant noise. Carrying a load on one side produced more veering to the opposite side. The detailed samples showed that, underlying the main trajectories, were alternating concave and convex (left and right) movements, suggesting stepwise changes in body position. It is argued that the same external and body-centred cues that contribute to reference-frame orientation for locomotion when they converge and concur, influence the direction of veering when the cues occur in isolation in environments that lack converging reference information.

What does not cause veering; what does not help

Kallie, Schrater, and Legge seem to be good at telling us what does not cause veering: inexperience walking without vision; and they tell us what does not help: an explicit indicator of intended walking direction and a skill for curvature detection. Here are some excerpts:

What causes this veering behavior? Our research suggests that a simple explanation, unperceived motor noise at the level of individual steps, may explain the veering behavior of blind pedestrians and sighted pedestrians who are blindfolded. (p. 183)
A pedestrian’s ability to walk a straight line depends on the availability and quality of sensory information about walking direction (Loomis & Beall, 1998; Philbeck, Loomis, & Beall, 1997; Rieser, Ashmead, Talor, & Youngquist, 1990) and on the capacity to execute movements in an intended direction. (p. 184)
Although blind walkers often use acoustic or tactile cues during walking, those cues are often uninformative and sometimes misleading. For that reason, we asked what factors limit straight-line walking in the absence of visual, acoustic, or tactile cues. (p. 184)
The lack of difference in veering behavior between sighted and blind participants suggests that a history of visual experience is not critical to performance on this task. (p. 187)
The results of the present experiment do not support the hypothesis that people who veer the least by the previously mentioned measures are most sensitive to path curvature. For emphasis of this
point, the participant (P3) who was best at curvature detection (threshold radius = 36.48 m) also was the one who veered the most in the straight-line-walking task (mean unsigned deviation 
1.73 m). (p. 188-189)
An explicit indicator of intended walking direction did not reduce veering behavior. In fact, in the case of the static perceptual pointer condition, veering actually increased compared with the physical alignment condition, in which there was no explicit pointer. The increased veer is due mainly to a larger linear component in the participants’ trajectories. (p. 191-192)

Veering? What veering? (and a new article)

I am posting these excerpts from Guth’s chapter as a reminder of the theoretical and experimental underpinnings of at least one of the variables we are dealing with in the WiiCane project:

Once the veering tendency of each of these individuals had been well documented, he or she participated in a series of about fifteen 20-trial training sessions. The training procedure was modeled after bandwidth feedback approaches used in athletic training (Schmidt, 1991). During each training trial, a participant started with his or her back to a portable wall, walked away from the wall, and attempted to stay within a 2-m wide, 20-m long simulated crosswalk defined by two parallel, ankle-level infrared beams. Whenever either beam was broken, the participant immediately heard—through earphones—
the direction of the error and the distance that had been traveled before veering out of the “crosswalk.” (p. 356)
All participants exhibited marked improvement over the course of this training and, as illustrated in Figure 18–1, the effects were still evident 5 months after the cessation of training. (p. 356).
While one participant reported a cognitive strategy that wasn’t consistently effective (i.e., “step to the left every few steps”), the others reported that they simply “learned what it felt like to walk straight.” (p. 357)
Our feedback system was unlike anything experienced during the everyday travel of blind pedestrians. It required that well-aligned participants attempt to generate a straight line path and stopped them when they had deviated 1 m to either side of the intended path. (p. 358)
Although blind pedestrians have many opportunities to experience straight-line walking, this experience occurs in the presence of continuous guidance. (p 358)
One of the surprising aspects of our training study was that the effects lasted for at least 5 months, the point at which we ceased taking follow-up data. (p. 359)
The training study revealed that practice with feedback is useful, but it remains unclear what elements of locomotion were modified during training. (p. 359)

Also, you may want to check out this other journal article:

Kallie, C. S., Schrater, P. R., & Legge, G. E. (2007). Variability in stepping direction explains the veering behavior of blind walkers. Journal of Experimental Psychology: Human Perception and Performance, 33(1), 183-200.

I cannot attach it here but will email the article to stakeholders.

Wednesday, April 8, 2009

light strip proof-of-concept


On April 7th I completed construction and testing of the proof-of-concept WiiCane light strip. The purpose was to test the feasibility of using the WiiCane setup with infrared LEDs embedded in a strip on the floor. The light strip provides waypoint markers which can be detected by the remote and software to provide additional positional and movement data. The active LED strip is an improvement on and supercedes the previous concept of using reflective strips on the floor to provide additional movement data.

This version was built as a six-foot strip, with infrared LEDs spaced 12 inches apart. The LEDs are constantly illuminated, drawing approximately 50 mA each and supplying 7 mW of light with a wavelength of 940 nm. The LED package chosen emits with a 34 degree viewing angle.

Initial results look very good - the remote is able to see the LEDs at a fairly oblique angle, down to about 50 degrees off-axis. The lights show up at a distance of at least two or three feet, even against moderately-bright background lighting. The bottom picture shows the view reported by the remote when two LEDs are in the remote's field of view.