optimum criteria for subject selection for testing

Steve,

I understand you will be visiting some of the sites and meeting with key site personnel.

I suggest you share the criteria below with the administrators and O&M folks so we will have the best chance of success. I hope others respond to this with other ideas or corrections.

Criteria:
- the subject should have the capacity to tap a cane back and forth as they move forward
- the subject should understands the concepts of left and right; or, the subject can move left or right depending on which ear he/she hears a sound.

This is a short list but I think if you present this to the sites they will know which children are appropriate and which are not. We might also ask that teachers work on these skills with the children in anticipation of the testing.I am guessing we can assume we will not get 8 kids at each site that can do these things, but the more the better.

I suspect others might be able to add qualitatively and quantitatively to this post.

posts from recent emails

Steve requested I place some of our emails here on the blog. Here they are.

- - -

Sorry I've been absent from this discussion all day as I have been out of town. I have thought more about this and I don't want to underestimate the abilities of the young children in this project. I don't know these students so maybe they are not as involved as the students I am accustomed to working with over the years. And are we talking 3, 4, or 5 year olds? There is a BIG difference in even 6 months in regards to cognitive and conceptual development.

I can agree with trying using the feedback in the left ear to see if the child will move more toward this side etc. Young children are typically adept at moving toward sound sources. So this might work for a child who doesn't really have a deep understanding of laterality. It might result in more of a move toward the direction of the sound in my ear type behavior, instead of a conceptual understanding of straight line travel. But if using the Wii feedback system illicits the desired behavior, this is a good thing.

In response to Gene's comment - Here's my big problem with this and maybe Bonnie or you could explain so my simple mind can get it . . . what's the difference between learning the word RIGHT and what it means, from learning a particular sound and what it means. They seem like very similar cognitive tasks. Both use the auditory pathways, both require the child to understand the difference between lef/right, and both require the child to pair a label (sound or word) with the proper concept and react to it. So what's the difference and why would one be better/easier than the other? In my gut I would suspect the word would be easier. But what the heck do I know!?

I understand this and am I don't disagree. What I am questioning is the conceptual understanding of left/right in children 4 and under.

I need to seek out literature on children learning the concepts of laterality. It has been my experience that blind children, even as young as age 4, can respond correctly to "touch your right ear", "touch your left foot", etc but they don't truly understand more advanced concepts such as "move your cane tip to the right", "turn your body left", until months or a year after they understand they have two of some body parts, and a left and a right side. You can see this development when you administer the Cratty & Sams Body Image for Blind Children checklist. Maybe we should ask the O&M specialist to do this with each child before data collection. It only takes about 20 minutes per child. This starts with body parts, then body planes, laterality, manipulating objects in relation to body to demonstrate spatial understanding of laterality, movements to left/rigth and directionality (left/right on others). I'll dig this instrument up and attach a file tomorrow.

Finally, what percentage of the subjects are children age 5 and under? If what we develop works for school age children and adults, then the project has accomplished some of it goals, right?

Bonnie

- - -

Hi everyone! I just saw a video by Dr. Bil Hawkins who used a WHISTLE to give feedback for veering while crossing the street. Now, I would never, ever suggest that we distract students while crossing streets by sending them coded messages about their veering (perfect scenario for the need for isolated veering training OFF THE STREET with the wiicane!), but this auditory feedback might be something that could work with Wiicane. We can ask Bil for details, but from the video it seemed that if the student veered a little to the left, he got a long, slow descending whistle. If he veered sharply to the right, he got a quick rising whistle. Bonnie, do you think kids who can't remember L-R could learn that if the "left veer" (descending) whistle was played in the left ear it meant they had veered to the left, and if the "right veer" (rising) whistle was played in the right ear, they had veeered to the right? P.S. I tried to log into the blog to publish my comments, it said it was emailing me my new password but I've gotten no message. So I'll continue to email comments, hope that's okay. -- Dona

project status and ongoing discussions

First, I need to apologize for not participating in the ongoing discussions that are occurring on email regarding which feedback types are appropriate for young children and also which style of cane use our system will train students for.  As I mentioned, last week  for me was consumed with preparing for and then attending a great conference at the Metropolitan Museum.  Now that's over and my full attention is shifting onto WiiCane for the next two weeks as we get ready to do our first field tests.

I need to reiterate that our group discussions must take place on this blog, so that we end up with a coherent record of our conversations, and so that all project participants can easily keep abreast of our developments and make contributions as needed. I understand that there were some oversights in distributing invitiations to  blog participants, and I also know that it can seem easier or more satisfying to just fire off an email to one or two people, but it is important to the overall project that we use the blog. I expect traffic to pick up significantly now, so it's really a very good idea to subscribe to the email digest that will come to you on every day when there is some blog activity. If you have not signed up for that yet, please do so by filling in your email address in the text box in the upper right corner of the blog and then clicking "subscribe".  If you have not received an invitation to participate in the blog, please let me or Gene know, and we will take care of it.

Regarding the question of how best to provide corrective messages to young children undergoing training with WiiCane: I think that this is a question that will be clarified quickly when we start running kids.  My guess is that directional audio that plays a tone in the ear on the side toward which the correction should be made will be more effective than saying "move right" or "move left".  But, it's easy to create a feedback system that allows for either option to be set, so we should have no difficulty in determining which works best, or, we provide both options and allow the test administrator to decide.

On the question of two-point tap vs. constant contact: my understanding is that most users these days are trained in constant contact, and so we need to ensure that our system reflects this preference. We should not develop a system that is only capable of providing instruction in an obsolete or unpopular method for cane use.  But, the proposal was written specifically to address two-point touch, and so we need to do that first. Once we get that working, we can augment the system so that it can also be used for training constant contact also. From a technical perspective, it is much simpler for us to first develop a tap-based system for motion capture. We know that tap events are very easy for the Wii's accelerometers to recognize. So, our current conception calls for taking a "snapshot" of the cane's position at the moment when a single tap occurs. If we define cane arc as the angular or lateral diplacement that occurs between tap events, our work is greatly simplified. Without taps, we will have to develop a much more complicated approach to arc width calculation, in which we would have to recognize the moment where the direction of cane movement reverses. That's not impossible to do, but it is probably a lot harder. My thinking is that we should master 2-point touch first, then introduce constant contact later.  The next couple of months are going to include a lot of trial and error as we develop our technology and our pedagogical approach. Nothing is fixed in stone right now, and we need to remain very open to making changes and enhancements suggested by our domain experts and our users. but, we should not abandon plans that were approved by the grant review panel without receiving explicit permission to do so. I also think that its sensible to start with something that we know how to do, get that working, and then move on to them more complex stuff.

On Tuesday, Zach and I will get together to connect his Java WiiCane drivers to the test administration program that I am working on. We should have something nice to show on Wednesday, when I am assuming that we will meet at the office to discuss our plans for responding to the IRB's questions about the test protocols we have requested them to certify.
sl

Progress report on application programming

Zach and I have made progress on the application that will run the WiiCane system.  This application consists of two components:

• The device driver, which Zach is programming in Java. This is a program that will run on the PC mounted to the pole at the end of the course. The driver establishes and manages the wireless connection between two Wii Remote units: one mounted on the cane and the other mounted on the user's body. The driver maintains information on: 
• Linear distance between the starting point and current location.  
• Body compass bearing, expressed as a positive or negative angle.  Pointing straight on the course returns a bearing of zero, and deviation to the left or right returns a negative or positive value, respectively.
• Amount of veering, measured linearly from the imaginary line connecting the start and finish points. 
• Cane taps events. A "tap threshold" is used to distinguish intentional taps from normal accelerometer fluctuations. Three taps in rapid sequence is a special signal that can be used to indicate that the user is about to start running the course.
• Wrist roll. As Gene mentioned, the system will not work if the user rotates his or her wrist such that the Wii remote mounted on the cane is not pointing up at the lights. We can assume that this has happened if the body-mounted Wii places the user in proper position on the course, but the cane-mounted unit reports that no lights are seen.
• Cane arc, expressed as a positive or negative angle of the cane in respect to the body at the moment when a tap occurs.
• The course set up and test administration program, carried out in Adobe Director, which I will be creating. We will use the Moka Xtra to act as a bridge between Zach's Java-based Wii Cane driver and Director.  We have tested this, and it works well, and is very fast, which means that we will be able to respond to user actions without perceptible latency, which would result in system sluggishness. The Director application that we will use for testing in November will include the following features: 
• A user log in screen so that subject parameters (cane height, body width, age) can be set prior to testing, and so that performance statistics can be saved to the computer's hard drive for later analysis.
• A course parameter screen that permits an administrator to customize feedback conditions.
• An animated plan view of the course with an avatar representing the test subject indicating instantaneous body and cane position and bearing during the test. 
• A training administration screen that shows elapsed time and a visual display of feedback prompts currently playing in the subject's headphones. 

These programs are being developed specifically for the upcoming tests. The actual commercial version of the WiiCane product will include other features and a more polished presentation.

wii cane feedback logic and testing procedures - draft

WiiCane procedures for initial testing

The following drafts are for everyone’s consideration and critique.

This weekend I visited Touch Graphics and saw the working model of the WiiCane and course. It is quite amazing. On the HD digital display at the end of the 32 foot course, you can see the cane movements precisely as the cane/subject moves. The equipment and remote software are working and it’s impressive. The user interface and the software to create the feedback loops are the next steps.

I will use the term “detectable” to mean that the WiiRemote can detect the infrared lights suspended above the course.

I will use the term “expected cane tip-placement” to mean that the cane tip lands within a tolerance of 2 inches on the left or right on the surface in front of the subject that is equal to the widest part of the subject’s body’s lateral plane.

These comments do not include procedures for deafblind subjects.

General considerations and concerns:
· The subject may not role their wrist as they move the cane; this will cause the cane to be undetectable. Subjects’ rolling behavior will need to be corrected with intervention.
· The subjects will need to stay with the detectable lateral area during testing. If they move out of this area, the tester will intervene to move then back into the detectable area.
· All subject will align for each trial on the course by squaring-off , placing the back of their footwear into cutouts of a wooden board properly located at the start of the course
· All subjects will wear customized goggles to occlude vision to the point of LP.
· All subjects will wear wireless headphones that are expected to deliver the feedback loop and prevent echolocation.


Logic for creating the feedback loops

VEERING
· When the subjects’ midline exceeds 16 inches from the intended straight-line of travel, the subject will hear a message
· The feedback loop will say the following is high quality electronic speech: “move to the left [or right]”
· The feedback message will repeat each time the subject exceeds16 inches from the intended path.

CANE COVERAGE
· The software will check the subject’s cane tip-placements after the first three lateral sweeps. Thereafter, the software will check the most recent three cane sweeps/tip-placement.
· Feedback will, of course, be issued only if the previous feedback statement has completed.
· The software will select a feedback statement based on the following logic:
- if the cane tip-placements for the three most recent sweeps result in the expected cane tip-placement, then the subject will hear: “Good”
- if the cane tip-placements for the three most recent sweeps result in cane tip-placement further from the subject’s midline than the expected cane tip-placement, then the subject will hear: “Too wide”
- if the cane tip-placements for the three most recent sweeps result in cane tip-placement nearer to the subject’s midline than the expected cane tip-placement, then the subject will hear: “Too narrow”
- if the cane tip-placements on the most recent three sweeps are asymmetrical (not too wide or too narrow on both sides), the greater lateral displacement from the expected cane tip-placement will determine the feedback and the subject will hear: “To wide [or narrow] on the right [or left]


Procedures for conducting trials

For the first trial for each subject, the subject will be have the experiment explained; the testers may use whatever appropriate language is necessary, and take as much time as they feel is reasonable for the subject to understand and be comfortable. Subjects will be asked if they are comfortable with the headphones and the goggles. Verbal and non-verbal response will indicate whether the subject should continue. They will then be walked/guided once down the course, shown the end of the course and the safety obstacle. They will then be walked/guided back to the start of the course. Subjects will always complete trials for veering before trials for coverage.

VEERING AND CANE COVERAGE

· The subject will align at the start of the course.
· The subject will be told to swing there cane and walk forward in a straight line
· The subject will walk unimpeded unless they become undetectable. If the subject becomes undetectable, the tester may physically prompt or manipulate him/her back onto the course from behind the subject.
· When the subject ends the course, they will be walked/guided to the beginning of the course.
· Each subject will complete 10 trials on any particular day, on up to 5 separate days. This number will be doubled at one test site. This procedure will be repeated for all trials for each behavior, veering and coverage.

Course construction complete

This picture shows the WiiCane experimental set up at Touch Graphics' office. The apparatus consists of two poles at either end of a 32' course.  A pair of overhead cables supports a plastic strip that houses wiring and electronics for 64 infrared LED's on 6" spacing.  We added the second cable after observing 8" of sag at the mid-point of the single cable set up; the second cable takes the shape of a catenary (like in a suspension bridge), and hangers drop from the upper cable to the lower one that carries the light strip (like the roadway in a bridge). This ensure that the light strip is flat, and that should simplify algorithms for determining position and cane movement. The second cable also appears to stabilize the light strip and to reduce sway from air currents. Immobilizing the lights is key to ensuring our system's measurement accuracy.

In the detail picture, we see the two cables at the start of the course, a support pole, and the plastic channel that holds the electronics. The underside of the channel is printed with a metric distance scale that will help us to calibrate the algorithm for calculating linear position on the course. To do this, we will mount a laser pointer on the body-worn Wii device, and (hopefully) we will be able to visually observe the red dot from the pointer when it appears on the measurement scale. It's not clear whether this scale will be of use in an actual Wiicane product.

We have tried out the apparatus in our office with a Wii device mounted to a cane, with very promising results: as we walked along the course, the system was able to easily see the lights.  We could watch the lights moving on the video screen we had mounted on the goal-end pole, and it seemed to work for cane swings that were quite wide. The only remaining problem, in my view, is that the user must not roll his or her wrist while running the course, because then the camera is not pointed up at the lights, and we lose position awareness. But, since our second (body-mounted) Wii device will always tell us where the traveler is in relation to the lights, we will know that  wrist rotation has occurred if the person is under the lights but the cane-mounted Wii doesn't show any lights visible. Then we can provide spoken feedback to encourage the user to straighten his or her wrist, and we can let them know when this has been done successfully before resuming travel.

discussion about pole placement in the coures

Bonnie responded to my last post via email, so I am pasting that  here, along with responses from Dona, Gene and Zach.  I want to make sure that our discussion is captured in the blog so that we end up with a nice record of our process.  The question of sagging has been taken care of by adding a second cable above the one holding the lights. think suspension bridge.  the upper cable takes the shape of the catenary, and the lower cable, the one holding the lights, is perfectly flat. 

(In physics and geometry, the catenary is the theoretical shape a hanging chain or cable will assume when supported at its ends and acted on only by its own weight. Its surface of revolution, the catenoid, is a minimal surface and will be the shape of a soap film bounded by two circles. The curve is the graph of the hyperbolic cosine function, which has a U-like shape, similar in appearance to a parabola.)  Click here for more information about the catenary.

Hi Steve,

I am just now reading the blogs about the overhead string of LED lights and the belt mounted Wii remote. Sorry it took me so long. As others have commented, getting rid of the path off the floor seems like a big improvement. I like the idea of the Wii remote being in the center of the user's back as it won't interfere with cane usage.  It sounds like you will be able to get more accurate feedback about cane arc with the overhead system.

 

Have you been able to set up the cable and poles and try it out yet? I'm curious about the length of a path, considering that sagging may lead to inaccurate measurements.  If you can't get anywhere near the 30 feet length, would you envision two strings of lights with 4 poles in order to cover the entire 30 foot path? 

 

Gene & Dona, what do you think would be the impact of having a pole in the path of travel and at the end of the path? Do you think any of the subjects would be able to use echolocation to hear the poles? I'm wondering if the poles could be used as cues for alignment in a way that might interfere with the veering data, which was designed to have no audible markers, so to speak. I read that the pole will be padded, which is good because I imagine some of the younger subjects or those who are not covering their body width bumping into a pole. Do you bumping into a pole would result in the person altering their gait or cane width as a response to contacting a pole? This is the real world. When my students bump into something because they were not covering the entire width of their body (or not paying attention, or cane hand not extended), they often will respond by swinging their cane wider. Some students will slow down because they are now aware there is an obstacle and they want to avoid contacting it. A slower speed and having to move around the pole could affect the subject's straight line of travel.

 

I'm trying to view this pole idea as a positive learning experience from an O&M perspective. If it is the best way to collect the data (cheaper, appartus readily available, easy to ship/assemble) then there just needs to be consideration of how the pole(s) might alter the subject's cane usage. There are obstacles in the environment so maybe the experimental design is now revised to include an obstacle. If it works out that we only need a pole at the beginning and end of the path, they my points may  not be relevant. I'm simply thinking aloud as to how a pole IN the path might alter the subject's cane usage and alignment.

 

Thanks.

Bonnie 

-----------------------------------

From Dona:

Wow, Bonnie, interesting points!  I hadn't thought about the possibility of the goal pole serving as an aid to aim for, or the effect of having an obstacle in the line of travel.  perhaps two things could be done to mitigate that:

1. Surrounding the pole with spongy cushiony material should make the echolocation very difficult (even with clicking );

2. Having a line run perpendicular to line of travel a few feet short of the pole will warn the user that he's reached the end.

-- Dona

--------------------------------

From Gene

Bonnie,

Subjects' vision and hearing are occluded.

A pole in the center of the course would not work. I believe that the 32 foot course will have a single suspended string of lights. The poles are anchored at the top and bottom.

The single forward pole at the end of the course should not interfere at all. It will be padded, of course, and subjects may encounter it when the course is completed.

Does that all make sense.

Gene 

---------------------------------------------

From Zach:

I was also skeptical about keeping the line taut but I've seen it at
Steven's office and it looks great.

The starting pole will have a wood block or similar on the pole or on
the floor to help subjects square off.  Both poles are wrapped in foam
to prevent injury.

---------------------------------------------

From Bonnie:

Gene - I obviously forgot about the subject's hearing being occluded. That sure eliminates my concern about echolocation. Yes, your comments make sense. I had thought they couldn't string the lights for the entire distance, hence a pole in the middle of the path would be needed. So my concern no longer apply!

Zach - I am glad that the line can be kept taut for the entire distance. This eliminates the pole in the middle. I'm excited that Steve has set it up and it works.

Dona - I like the idea of a textural change in the floor a few steps before the end of the path to let folks know they have reached the end and the pole is near.

Thanks.

Bonnie

belt-mounted Wii remotes

If it turns out that the overhead string of lights works well for tracking cane arc, I would also like to try a belt mounted unit to watch for veering. It may be that we don't need a camera at the end of the course if we could see at least two lights straight up.  Here are some things to consider:

1. if the wii device is mounted on the person's body instead of on the cane, we should be able to determine placement on the course (y-axis), veering (up to the point that lights are no longer visible), rotation in the x-y plane, cane centeredness (by comparing values from both cane and body-mounted sensors.

2. We have to figure out where on the subject's body to attach the wii device. I am thinking that putting it in a belt holster on the midline-rear of the subject's body, because that seems to me to be less likely to be blocked by swinging arms or upper body. But certainly, we will have to test that and it might be necessary to make a special belt that can hold the device at a precise angle.

sl

Overhead course pole design

Here's an sketch for an idea for making a wiiCane course with an overhead string of LED lights that will allow us to track position along a course and also to observe cane arc and veering. The best part about this is that it is mostly parts from IKEO, so it will be very inexpensive to set up a course. The picture shows a pole with a video screen and a plywood box mounted to it. At a point 8' above the floor, there's a fitting on the pole that is used to hold each of the three cables in tension. A cable comes out of the middle of the pole to supply electric and control to the lights.

new concept: putting the IR lights on a cable overhead

After thinking through the problems associated with constructing a floor mat system for tracking position of a student during cane travel training with a wii remote, I want to consider a new idea that may resolve many  problems.  This approach calls for creating an overhead string of LED lights on a three or four wire cable stretched tightly between two poles.  The poles would be floor-to-ceiling aluminum rods of the type I have here in the office to hold up the desks. These are telescoping poles from IKEA that extend up to 12'. I believe this could be easily installed in a wide variety of places. Even if they didn't have a full 30', they could use set up and use a shorter course.

Here are the pluses and minuses of the proposed system:

1. We can control the distance between the camera and the lights this way. In the floor mat arrangement,  the shortest cane held the Wii device only about 18" or so above the floor, so the portion of the floor that the camera could see at once was very small. If we point the wii device straight up instead of straight down, and if we string the lights on a single cable directly above the course, I believe that we should be able to be far enough from the lights that we would mostly always able to see one light, and usually more than that.  In the floor scenario, we were talking about making multiple rows of lights so that the cane would be seen even when the subject was not standing in the middle of the course. If we are far enough from the lights, we may only need one row. This has to be demonstrated experimentally.

2. The poles will be rigidly held in place, so, as long as they are padded, it won't pose too much of a hazard in the case of inevitable collisions. Because the poles are rigidly supported, we can draw the cables very tight, reducing the amount of sag. We will probably need some kind of mechanical tensioner in the cable for setting up the apparatus. While a little sag might not matter, a large amount of sagging could lead to measurement inaccuracies.  I think this set up is a fairly low hazard level.

3. This apparatus will be much easier to package and ship, much, much cheaper to produce and less prone to damage and wear, because no one will be walking on it.

4. The system will be very scalable. It will be relatively easy to extend or reduce the length of a course. It would also be possible to include additional legs, as long as the subject is prepared to negotiate a free-standing pole along the route.

5. There may be other applications for the set up that I am envisioning, such as Wii Fencing. I just googled it, and it does not appear that wiiFencing has been done. Here's a link to a thread on that subject. A cheap set up for tracking linear motion along a virtual course may have a range of uses in both therapeutic and gaming domains.

So...I am going to go shopping tomorrow and buy stuff to set this up. I will be ready to start testing next week, and if it looks promising, the next thing we have to do is to go to all of the sites and make sure that they have a location where the poles could be set up.

sl

proposed design for wiiCane modular walking surface

After this weekend's meetings, in which I met with Zach Eveland, Annette Gourgey, Raphael Baptista and Gene Bourquin to make plans leading up to our test version of the WiiCane apparatus to take place at four sites in November. We made significant progress in several directions, and I am hoping that Gene, Zach, Rafael, and Annette will also produce posts on this blog capturing their view of the work as it is right now.

I am going to discuss the design for a modular walking surface for the WiiCane system. We decided to go this way after considerable back and forth discussion. We decided in the end that the floor mounted lights will be easier to track along the entire distance of the walking course (compared to lights mounted 30' away at the goal end of the course. Also this approach gives us the ability to make any length track, and even to make courses with turns, etc. (although I can't think of a reason to have that yet). This approach will be more expensive, and much more difficult to transport and set up. But, there are also distinct advantages, and it will be much simpler to achieve reliable position awareness doing it this way. And there are additional complexities, for example because the IR LED's have some thickness and they are crushable, we have to cover them with a surface that will protect them and the associated wiring harness. We also have to camoflage them tactilely, so that cane will not discover them, thereby giving away the straight line travel path, which is what we are training our users to do.

I am proposing a series of folding lightweight panels. 5 of these would lie end-to-end, creating a surface that is 8' wide by 30' long. The panels will be extremely light, and when all folded up, they would fit in a box that is 6' by 4' x 6", which is very manageable. It would fit easily in a station wagon. I am thinking that the panels would be shipped in a wooden box that would then become a head unit at the goal end of the course. This would house a camera, flat panel display, speakers, computer and any other devices. Each of the five panels will consist of a 4' x 6' fixed section and two hinged side pieces that are 2' x 6'. The panels will be 3/8" thick masonite board. When the hinged panels are folded in (covering the treading surface), the whole thing is 3/4" thick. The treading surface will be a long permanent self-adhesive floor banner from the printer we use to make tactiles. that printer produces 30" wide prints of any length, so we will burnish the prints to the masonite and trim off the excess. I want to try to use the prints themselves as a continuous hinge for folding up the panels for shipping, but that will take some experimentation.

The printing on the treading surface can include a long ruler in metric and english units, a big wiiCane logo, and different color lines to indicate set ups for various courses (green course, red course, blue course, etc). I just remembered that I once designed a book cover with Elga Joffee that we nicknamed the Corridor of Pain. It was a rendering of a theoretical hallway with one of each of the hazards disallowed in the ADA regulations, things that you can smack your head on, etc. I was thinking that we could base the placement of obstacles in one of the training course on some of those things, as a way of preparing for those kinds of conditions. The obstacles that we actually use would be inflatable or Nerf material. I like the idea for a future line of nerf obstacles! In any case, we have to start thinking about the pedagogical objectives and capabilities of this way of training people. We should end up the project with a producct manual that explains our reasoning and sources. We should also all be starting to think about publishing this work soon, assuming that we get some good outcomes.

The lights will be in three strips that will run the entire length of the course, one right down the middle and one in either of the two wings. These will be marked on the printed floor surface, so we can just use that as a template when drilling holes. We would pop the little LED through a hole so that they are slightly recessed from the surface of the treading surface, then fill the holes with clear epoxy. This will level the surface and protect the LEDs while allowing IR light to pass through. A cable will connect the lights to a control device in the head unit/storage crate. I don't yet know how to join the units to one another mechanically or electrically yet, so any ideas about that would be appreciated. A big neoprene rubber starting block would be at the opposite end of the course. This would have little semicircular cutouts for their heels while squaring off.

So, please give me feedback. I am going to start building a sample on Wedenday, with the hope of bringing it up with me to Boston during the week of September 28.

sl

wiiCane version 3.0

This new version of the wiiCane is a significant improvement over the previous one. The construction is similar to version 2.0 (machined and welded aluminum), but it is smaller and lighter. the aluminum fixture is very rigid, so you can really feel every nuance of tactile information transmitted up the shaft of the cane. It also seems to be well-balanced, regardless of how you point the wii remote. So, with this design we can experiment with a variety of light placements, including at the goal end of the course, on the floor, on the ceiling, or on the user's body. We now have to decide whether to use the adjustable length cane, or if we should just make four of fived fixed length versions. If everyone agrees that this is a good approach, I will ask Accufab up in Ithaca to make five more, and we can experiment with various finishes and branding strategies (logos, etc.)

WiiCane version 2.0

This version of the WiiCane is made by chopping standard graphite shaft cane and removing an 8" section near the grip. We then installed a custom aluminum fabrication that returns the cane to its original length. the aluminum part holds a pivoting cradle, which in turn holds the Wii Remote device. The device can be pivoted around 360 degrees without significantly unbalancing the cane so that we can test using IR light sources in various positions. In terms of balance, stiffness and general feel, this version is considerably better than the previous plastic one. The tradeoff here was weight. This version is significantly heavier than version 1.0, and so we have re-engineered the part. It will arrive this week from Accufab in Ithaca. I will put together a new prototype for testing and see if the aluminum version will work. The next cane-based question is whether we should use multiple fixed-length canes or if we should try using an adjustable length cane. Gene is looking into that question now.

Research on veering

According to an article in the New York Times, our work on WiiCane is highly relevant to ongoing research in the multi-sensory perception. The article discusses work by Dr. Jan L. Souman of Max Planck Institute for Biological Cybernetics. Dr. Souman is seeking scientific explanations for the "lost hiker" phenomenon, where people without access to visual cues always loop back on themselves when they are intending to walk in a straight line. One interesting outcome of this work is that it debunks the theory that the tendency to veer is an outcome of one leg being shorter than the other, since circling can happen in either direction (in the same individual). While WiiCane is not intended as a training system for hikers, I think we should pay attention to this parallel line of inquiry. If we can demonstrate that, using corrective and supportive audio and/or vibratory feedback, it is possible to teach people to walk straight, the research community may be interested to know about this. I sent Dr. Souman an email about WiiCane, I will post his response if he replies.

WiiCane Course Layout-Scheme No. 1

This image shows one way that a course could be constructed for the WiiCane. Here, a 30' path is marked on the floor. A user is shown walking along the path towards the light bar at the goal end. The light bar is an 8' long horizontal beam that is supported on two adjustable stancions. In this rendering, the light bar is about 40" above the floor. A visual spectrum camera is mounted to the mid point of the light bar. This camera faces the user, and is intended to track a color dot that is pinned on the person's torso. The camera will track the person's absolute x-axis position (left or right of the travel path). Then, lights mounted on the face of the light bar provide information on arc width, arc-centeredness, and arc height at mid-swing. In this version, there are no lights on the floor. In this image, the floor has visual markings to help an observer determine the user's position in terms of distance from the starting point and amount of veering. A 24 inch wide starting block allows the user to square off at the beginning of the route.

wiiCane version 1.0

This image shows a photograph of our first wiiCane. This is a standard mobility cane that has been modified as follows:

1. An 8" section (just below the grip) of the cane was chopped out.

2. A plastic part was inserted in place of the removed section of the cane. This part looks like two tuning forks connected end-to-end at their tines.

3. A smaller 3d printed part (painted red here) was mounted with screws to the first part so that it can pivot 360 degrees. This smaller part holds the wii device.

The completed wiiCane is slightly heavier than a standard cane, but not so much so that it becomes difficult or awkward to use. It also appears to be well-balanced, regardless of how the wii is rotated in the fixture. Now that we have this assembly, we can begin to do in-depth testing of a variety of different light configurations, so that we can determine which arrangement offers the most options for motion tracking during cane travel exercises.

Here's a picture of our new 3d printer. It's the UPrint desktop printer from Stratysis. I got it for the purpose of outputting prototype enclosures for the NimblePad that we are working on under contract to Nimble Assessment Systems. But I am hoping to use this for many other things, including making lots of audio-tactile models and other touch-sensitive interactive displays. With the new usb sensor device that Zach is making producing now, and plastic models printed from this machine, we can make detailed very strong exhibits cheaply and rapidly. Right now, I am printing a pivoting fixture for mounting the wii remote to a mobility cane. I am hoping that other interesting new things will emerge from this, and would welcome comments or suggestions.

steve

wiiCane curriculum pedigogical objectives

Now that we are starting to get more detailed in our understanding of what can be accomplished with the envisioned WiiCane system, we need to start thinking also about how it will be most effectively used in actual teaching. While it is too early to make any decisions about specific lessons or activities, we would be smart to begin developing how instruction with the wiiCane used in a rehab or school setting will probably occur. Gene will be the leader for this part of the work, and I am hoping that he will faciliate a discussion about teaching parallel to the one we are having about the technical aspects. We should start brainstorming about how we image using the system in a teaching environment, because that discussion will shape our decisions about what to emphasize in taking measurements.

As a starting point for this discussion, I suggest that we look at SAL, which is Sally Mangold's curriculum for braille literacy. In that, there are a set of short actitivies that are organized in order of ascending difficulty and skill mastery requirements. I am not sure if the aquisition of travel skills is analogous to learning braille, so it may be the wrong model, but this gives people something to respond to. Gene?

Mobility perspectives for post-Kalamazoo

For technical reasons please continue by reading the follow up comments to this post.
Thanks Gene

Technology options

The following is a list of the technology options we are considering for obtaining information about a traveler's movement and their cane handling technique.

• Inertial measurement - uses movement information supplied by the Wii Remote's built-in accelerometers (and possibly the Wii MotionPlus gyrometers). This gives relative movement information from inside the Wii Remote's frame of reference.




• Good for: detecting taps, determining cane orientation, detecting cane movement in the absence of other sensing methods


• OK for: measuring arc width


• Not good for: determining cane position, detecting veering






• Light strip - an array of computer-controlled infrared lights on the floor which are visible to the Wii Remote's IR sensor. Like the light bar method this gives us an idea of the traveler's absolute position.




• Good for: very precise arc width measurement when lights are in view, precisely determining traveler location, detecting veering


• OK for: might be able to measure cane-centeredness and coverage


• Not good for: easy installation (requires careful positioning of light strips or installation of walking surface)






• Light bar - similar to the Wii's Sensor Bar this method uses the same principle as the light strip method - an array of infrared lights which can be seen by the Wii Remote's IR sensor gives us an absolute position reference to work with. In this case, several lights are mounted horizontally at the end of the course and are seen by the Remote's IR sensor.




• Good for: arc width measurement, detecting veering, determining traveler location, simple setup


• OK for: might be able to measure cane-centeredness and coverage


• Not good for: less accurate than light strip






• Rear-facing IR sensor - a second Remote or IR sensor mounted on the cane is positioned to view the traveler. Markers or lights affixed to the traveler's clothing are seen by the Remote or sensor.




• Good for: very precise cane-centeredness and wrist isolation measurement, precise measurement of coverage if arc width is known


• OK for: might be able to measure arc width


• Not good for: detecting veering, determining traveler position






• External camera - a computer-connected video camera at the end of the course records the traveler's progress. Computer vision software analyzes the video to determine information about traveler and cane movement.




• Good for: measuring cane-centeredness and wrist isolation, detecting veering, measuring coverage if arc width is known, easy installation


• Not good for: determining traveler location, precise measurement of arc width or coverage, long courses - even a 30-foot course may be too long

As no single method provides all necessary information, it is very likely that the ultimate embodiment of the WiiCane system will use two or more complementary methods. We will continue to refine the inertial measurement methods and investigate the other options.