This past weekend I drove down to Auburn to help with installing the LipSync for my first user, Mr. M! It was busier than expected, as I accidentally scheduled the trip for a home game of Auburn vs. LSU.
I had pretty minimal coordination with Cyndy, my contact at ALSA, mostly due to my lack of time outside of work. But we were able to set a date and I got up Saturday, packed my car, printed some directions to Auburn, and hit the road! The drive down took about 4 hours, and then I spent an additional hour lost inside campus, just trying to find the right neighborhood among the teeming masses of people present for the football game.
Finally I found the house. Cyndy was already there so we unpacked and got to work. The main order of business was figuring out how to mount the LipSync to Mr. M's wheelchair, which proved surprisingly difficult.
The first candidate was Loc-Line, a new (to us) product Cyndy had found which is a snake of linked plastic segments that snap together and create a fairly rigid arm.
In true "if you have a hammer, everything looks like a nail" fashion, I printed a few pieces to test making an adapter from loc-line to the LipSync, only to discover that the weight of the LipSync was too much for the length of loc-line required. A LipSync can be supported by about one foot of the stuff, any more and just the weight of the mouse will drag it out of position.
We worked with a few other mounting ideas before settling on a flexible webcam arm purchased online. It had less than ideal clamps to mount to the wheelchair, but would work decently well on a desk. Finally, Mr. M was able to take the LipSync for a test drive!
I didn't take very many pictures, as I was essentially a stranger in Mr. M's home, but I did write down notes from the trip so that Cyndy and I can iterate and improve on the experience.
- We didn't have a good solution for mounting the LipSync. In my testing I had just held the LipSync... With my hands. Pretty glaring omission on my part.
- The LipSync does have a learning curve, and I need to find the best way of training someone new on using it to minimize frustration.
- The LipSync needed some effort in configuration. Windows 10's default mouse acceleration profile was bonkers, and we had to tone it down to make the mouse usable.
- The mouthpiece filter needs to have the extra locking piece. It popped off the LipSync within the first few minutes of use. We wound up supergluing that one together, with a promise that I'd fix it properly later.
- The LipSync tended to move the cursor while Mr. M was trying to click. Perhaps a firmware change - one that locks out the first second of movement in case the user is trying to click - would help.
- The i3Mk3 survived 4 hours in the car! I was able to drag it out of the passenger seat, set it up, and start printing pretty much straightaway (after a Z-Calibration). My old printer would not have handled travelling so gracefully.
- We were able to get everything together enough that Mr. M was using a LipSync! There was a reasonable chance that we'd be stuck at some hurdle and have to leave without reaching that milestone, I'm so glad we made it!
- After trying out the LipSync, Mr. M's testimonial is that it was easier to use than an eye tracker. That's a pretty positive review!
Takeaways and future improvements:
- Cyndy has ideas for more AT devices we can print. I'm very excited to help with developing those!
- Having a wireless LipSync variant is the future. I need to investigate if the next gen Adafruit "Feather" based LipSync will have bluetooth.
- When installing a LipSync, it needs to be 100%. There's not really a "eh, good enough" use case, because of the difficulty involved with just using the device to begin with. I need to get the mounting and user experience (software setup) really tuned in, otherwise the LipSync on its own isn't useful!