It’s been two weeks since my last post, and a great deal has happened since then. Most notably, the second version of the handstand has been constructed and set up. Take a look at the video below for a quick demonstration of that and the new and improved debug program (The thing graphing data on my laptop). In the video I demonstrate the new debug program screen, the hand opening and closing with the EMG system, and all of the hand functions that are currently loaded: pinch, peace sign, rock sign, and two types of finger wave. Most of the functions are placeholders, but they’re still fun to see.
The weekend before last, I went to the FRC world championship in St. Louis. Immediately before leaving, I began to laser cut new parts out of acrylic for a second “handstand”. Here are some pictures of that process!
While I was in St. Louis, I either spent my free time eating sandwiches at this awesome place called Snarfs, or writing code for the new hand test platform and the debug program. I was able to make some decent code for the hand, and some less than decent code for the debug program. When I got back to PA, I encountered several bugs with the code, but the major ones were all quickly squashed to produce the test platform that you saw in that video. The new code has a lot of new features, and will be able to be easily converted from servo control code into regular motor control code. It can be found in the GitHub repository under multiEMGServoTest. Aside from the averaging code needing optimization, I think that the program is in a really good place. I’m fairly certain that at least some of it will end up in the final version of the hand!
The new features include preset hand functions (as seen in the video), EMG data averaging (to ignore random EMG spikes), code to ensure that the EMG value is not too high, and code for hand “commands” – commands that recall presets or reset certain threshold variables. The debug program can graph in multiple colors for different EMGs, and even has some rudimentary code in place to send those aforementioned commands to the hand. Pretty cool stuff, overall!
We got a chance to test out the Handstand v2.0 at the Philadelphia Science Carnival just two days ago, on Saturday. The hand was a huge hit, and hundreds of kids got to try activating it! (Probably more than 99% of them were able to use it) We arrived at the event around 7am, and got back to Westtown at around 8pm, after a long but rewarding day in front of the Franklin Institute. I have to give a huge shoutout to Alex Nunes and Casey Dech, two other members of the hand project group, as well as Tom Barnett (A helpful Sophomore), for taking some of the long shifts at the science carnival and helping with the incredibly tedious and frustrating task of stringing the servos and the fingers together. Being able to demonstrate the hand to an audience was a nice new experience, and I’m glad that we were able to get everything working in time for the event.
As this week progresses, I plan on testing more of the EMG averaging code (which already seems to work pretty well), and testing the use of multiple EMG readers at the same time. I’ll also be designing a test for the hand recipient. On Friday, there will be a meeting between myself, other members of the hand group, and the hand recipient. This meeting will show us whether or not the hand recipient can use the EMG system that has been developed so far, and will be an extremely valuable test of the control system. I’m not too worried about the tests going badly – as I’ve probably said before, even if we can’t get a reading on his arm, we can still use other muscles like biceps and shoulders for EMG activations. This is a really nice control system, and I’m almost 100% certain that it will be a precursor to the final control system that we’ll end up using.
On that note, I’ll end this post. My next blog post will probably focus on more of the technical side of things, mainly about how the programs work at a basic level.
Until next time!