Arduinos are like miniature programmable computers that utilize both hardware and software to conduct a given logic-based task. It is composed of a motherboard-like base structure that can be wired with sensors, lights, speakers, etc. to conduct a given task. Any written code can be linked to this board to manipulate the outcome.
The Workshop & My Results
Our first workshop involved working with arduinos to conduct mini tasks such as blinking a light or playing a sound. We used a pretty inexpensive arduino hardware and software package from Adafruit and followed an instruction manual to both wire and code the task.
The first task that I completed was making this blue light blink.
This was honestly quite easy since all I had to do was follow the diagram to circuit the board, and then copy the code onto the software. I will admit, my background in engineering (with both circuitry and coding) made this a breeze.
I then tried playing a beeping sound with the instructions given in the packet. Once I did this, I tried to play around with the code to play a song. Here is my attempt at Hot Cross Buns:
It was a bit hard to figure out how many spaces to put between notes and what notes correspond to what sound, but you can kind of hear the tune which I'll happily take as an achievement.
Reflection
I really enjoyed working with the arduinos. I will admit, I'm not 100% fond of hardware/software, but it is definitely something I should learn. Also, it's something that will inevitably be a part of my career, so the more exposure I get to it now, the better. I also appreciate that the Adafruit package makes it easy to use and understand. However, there seems to be a big learning curve in the product which excites me in the possibilities of this product, but frustrates me right now as a novice.
This also made me think a lot about Daniel Canogar's work, and how he has a computer engineer collaborating with him to create his digital data-centered pieces. There must be a lot of software and hardware utilized in his pieces that work similarly to these arduinos.
Applications
If you want to program a device to do something, it probably has some type of arduino-like structure. I decided to focus on one specific project that relates to my personal interests for my research. Naturally, I took to Reddit to find different arduino projects. (Side note: I say naturally because doesn't this just seem like a Reddit-like topic? I don't know how to explain it...) Surely, after a literal 2 minutes on Reddit, I found a project of interest:
"How to Use EMG Muscle Sensor to Control Anything (MyoWare) - So with this sensor you can control a robotic hand, or you can use your muscles to control anything."
Here was the video linked:
Being in a prosthetics-design club, this really interested me. Right now, we're working with a little boy who does not have extremities past his elbows and knees, so we are designing cheap 3D printed prosthetics for him to complete tasks such as throwing a football or holding a cup of water. One issue we try to overcome (for cost and ease) is making something that easily clips onto whatever object he wants to hold. This makes the product purely mechanical (as opposed to electrical). When conducting research on prosthetic arms produced commercially, there are many cool products such as the DynamicArm which is a myoelectrically-controlled elbow joint. A video of this product in use is shown below.
As you can see, Sherri is able to easily open and close her prosthetic hand through the myoelectronic sensors. This would be amazing to implement in our designs, and truthfully, it always seemed like a distant idea until I stumbled upon that Reddit article.
Arduinos are relatively cheap and the MyoWare muscle sensor is only about $40. This opens the door of possibility to utilize such technology in our prosthetic design.
There are just a couple of issues...
Though the arduino board isn't that big, it's still too big to have attached to little boy's arm with comfort. Also, reattaching these sensors each time seem tedious. There were also some worrisome quotes on the Reddit thread about how well this technology works. Here's one:
"Hey, so I did my bachelor's thesis, not with this but with the MYO armband, which basically has 8 of these emg sensors in it. It f**ing sucked, the readings were really unreliable and I had to do a lot of data processing to get some amount of gesture classification to work. Also sweat and temperature changed readings over the course of a session."
I did find some sweat sensors that could be utilized to help with this problem. It's also important to note that they used an armband, not the mini sensors, which could have changed things. I do like the idea of an armband better though since it alleviates the problem of reapplying the sensors, so it's sad that it didn't yield them good results.
Overall, I think I want to try to utilize these sensors with the arduino board to maybe make a small muscle-controlled system that I can tinker with. This will give me a good introduction into the limitations of the product, and help me better understand how to incorporate this sensor into prosthetic design.
コメント