Capturing the Rhythm of Knitting

Smart Needles

I've spent the past few weeks prototyping how to capture the movement of knitting. My initial idea was to embed the hardware into the knitting needles. I started by attaching the components directly to my needles. I used an Adafruit Feather and an MPU-6050 accelerometer and gyro. I knit with the needles and sent the data from the MPU-6050 to MaxMSP (a program for interfacing with media).

This was my first time using Max, so I had a lot to learn. I attended a MaxMSP Jitter Club meeting and met with some classmates who showed me some basics, and T.K. helped me find a video tutorial for creating sound with sensors.

I was able to get started making sounds pretty quickly, but they were not very pleasant. I am still with making sense of the data. Another issue was that most of the movement was with my right hand, but the needles needed to be able to alternate positions because the yarn passes back and forth between the needles.


I realized I could solve the alternating needles problem if I started detecting the movement in the hand. I really like working with soft circuitry and the design challenge of wearables, so this seemed like a really good route.

Inspiration for my prototype

I made a glove out of white Neoprene fabric. I sewed the Flora on the back of the wrist and the accelerometer and gyro board on the back of the hand where it would move the most during knitting. While observing my own knitting I discovered that the hand moves much more than the wrist and would be the ideal location. Placing the components on the back of the hand would keep them out of the way for knitting.

I was really happy with how my glove turned out, but I ran into an issue where the Flora would suddenly stop sending numbers. I figured it must be an issue with my connections, so I used glue to insulate them and keep them from fraying. The problem only got worse, but I discovered that the short was caused by the Flora moving around. If I left the glove still the data would keep coming. While the accelerometer needs to be in the place with the most movement, the Flora needs to be in the place with the least movement. I will need to rethink this design quickly.