Live Knitter

Interactive Installation


In Greek Mythology, three goddess, known as the Fates, control the thread of life of every mortal from birth to death. In this modern age, data could be considered the thread of one’s life. Collecting an individual’s data on the mundane action they perform, one could weave a tapestry representing their live, from what websites they browse to which subway they take. Live knitter is a data-driven installation that brings together data, life and thread. It converts the real-time data of the world-wide birth and death statistics into knitted fabrics.

What viewers encounter first is a stacked bar chart updated in real time. The whole width of each horizontal bar represents the number of  births in one minute. The black segment represents deaths, thus the white segment amounts to net growth. A complete visualization image takes 24 hours to generate, ending up with 1440-pixel height-the number of total minutes in one day.  Consequently, the installation takes the same time to knit the pattern accordingly. It works in a fully automatic fashion, symbolizing that there should be no human intervention in the natural process of the life cycle .

If data is the thread of life, who takes the goddesses’ place to spin the thread? The average person may be unaware when governments and companies are widely collecting personal information. Nonetheless, their decisions based on the data would exert real impact on our lives and society. Live knitter serves an analogy to the data-based decision-making cycle in reality. When transforming data to fabrics, it turns the digital into physical objects, and entitles the invisible to a tangible form. The project strives to raise viewers’ attention to the close relationship between data and life in an artistic practice.

Documentation Video


  • Final Presentation.001
    In ancient Greek, people believe that the Fates, three goddess, controlled the thread of life of every mortal from birth to death. In this modern age, data could be considered the thread of one’s life. Imagine that you got all the data for one person, what do you see? You see the person’s entire life, from what he likes on social media, to how often he takes subway. The prediction from Black Mirror could become reality.
  • Final Presentation.002
    My capstone brings together data, life and thread. It’s called Live Knitter. So what is that? To put it simply, it is a data-driven installation that converts the realtime worldwide data of births and deaths into fabrics, using a knitting machine.
  • Final Presentation.003
    When I started to develop my Capstone, I immediately identified two questions what data and what form? I would like the data to be real-time updated and demographical. By demographical I mean that it is generated by peoples’ mundane actions, rather than by nature. Social media data is a good example. However, not everyone uses the same platform. So I turned to statistics of worldwide births and deaths because the topic is universally relatable to everyone.
  • Final Presentation.004
    This project is technically based on one of them called AYAB. AYAB allows a knitting machine to receive patterns from a computer. Without it, a machine only supports built-in patterns or patterns marked by punch cards. AYAB provides the schematic of a customized Arduino shield. The hacking is done by reconnect the machines wires to Arduino with the shield.
  • Final Presentation.005
    The data visualization in this project is stacked bar chart,, generated by Python Matplotlib. Each horizontal bar reprints the number of births and deaths in one minute. The total width is the number of births, and the the black bar is the number of deaths. Subtracting later one from the other, the white bar is the number of net birth in a minute. When the software is running, it is generated in realtime. Therefore, a complete visualization image that shows the data of a whole day has 1440 bars, which is number of minutes in a day.
  • Final Presentation.006
    For each bar, the knitting machine interprets it as an image of size 1* 100 pixels in the form of a binary array of 0 and 1. Following the binary data, then machine then pushes the needles on the flatbed to different positions.
  • Final Presentation.007
    When the carriage passes by, it threads the needles with different color of yarns according to their positions.
  • Final Presentation.008
    My project does not copy AYAB unmodified. Remember that in order to knit in realtime, pixels are generated every minute, adding to the pattern which the machine is knitting. I tailored the software by identifying the key functions that handle the image and communication, and cutting off the GUI module and the rest. Also it is necessary to change part of the settings in Arduino firmware.
  • Final Presentation.009
    Most knitting machines in the market are for home use. They need people to move the carriage back and forth. Live Knitter turns the machine into a fully automatic fashion because the process of birth and death should not be interfered by human force. Live Knitter automatizes the machine by tying the carriage to the slider on a linear guide rail, and controlling the rail with another Arduino.
  • Final Presentation.010
    As my capstone finished, there raises another question. If data is the thread of life in this age, who takes the goddesses’ place to control the thread? Governments and companies are wildly collecting data these days. We might not be aware when gathered data from. Nonetheless, the decisions made by government and companies based on the our data would eventually have an actual impact on us, just like Live Knitter is transform something invisible to something tangible, something in digital world to something in reality.
  • Final Presentation.011
    During the development, the feedback from user testing helped me decide what kind of data visualization to use. I started with the Heatmap on the left which represents data in seconds. I’ve also tried the stacked area. The heat map easily makes users lost because it seems too random. What’s more, heat map causes aesthetic fatigue because a lot of artworks have implemented it. An interactive artwork never completes. In the future, I would explore usage of fabric patterns in data visualization, which adds to the richness to the surface, and differentiates this physical representation on of data from the one on the screen.
  • Final Presentation.012
    A more serious problem is that the data I used for now is fake, The software now randomly generates data, instead of requesting from the data api. The ironic part of this project is that fake data does not visually or conceptually influence the output because later I figured out that the data api, though titled worldwide births and death, is not real it self. It is a simulator based on the current population and birth rate. Although we’ve heard of the age of Big data for many times, we might underestimate the difficulty for government and organizations to put together data from different sources, not to mention their willingness.
  • Final Presentation.013
    credit to AYAB, to Claire William an artist who does remarkable researches into textile and digital images, and to the book Knitting Technology
  • Final Presentation.014