A few months ago I submitted the idea to the group as part of the regional project for Burning Man, for a game that could measure “love” between two people. By using two galvanic skin response (GSR) modules attached to an arduino, a python script pulls the values of each module and updates a JSON file when the values change. The goal is to have two people stare into eachothers eyes similar to what is described here, while we measure their physical response and provide points-based analytics to the participants. Based on the distance to the two scores, we can determine how in “Love” two people are in or are falling into. To better sell the concept, the GSR was described as measuring frequencies. This made the Love Machine more palatable for most participants, albeit a little bit misleading.
The construction started with the code, after a few iterations as the idea developed into a full blown analytics platform. Starting with basic proofs-of-concepts (POC) that worked to accomplish the most basic tasks that were required to make a functioning system. Over the course of N-number of commits, I made the the app work in a way that anyone wandering through the desert could understand. When it came to the physical construction I enlisted the help of a local steampunk legend. With his help we were able to collaborate and build a sort of a kiosk that had a classic arcade look and feel to it. Once the physical enclosure was created, getting the Raspberry Pi to work was the next task.
As for what went really good, the general flow of getting the a lot of work done in my spare time made the delivery of the project not feel rushed. Having a clear vision of how to implement a stack that could accomplish the proposed task was a huge boost in confidence, as the logical way to structure things just seemed to make sense.
With most of the design being done on a regular laptop, a last minute redesign was needed when the constraints of the machine required the monitor to be turned on its side. This reduction in the screen width made the creative positioning of the elements more modern looking. Making a Raspberry Pi do what you want is not always that easy, the biggest issue I faced was getting a web server up and running, Node did not want to play on a Raspberry Pi model-A due to the limited amount of memory. Rather than just fight with Raspbian on a model-A, I switched it up and went with UbuntuMATE on the model-B plus. Rotating the screen is a simple setting in the config.txt.
Once I got out the the desert with this project I realized I still was not saving the data and that the Python serial library was not installed. The code needed to save the data to local storage was accomplished in an hour on the Playa, with no Googling! After the Python script failed to initialize I had to think fast, while there is wifi out there there is not a lot of wired internet. I was lucky to have already of networked on the Playa with someone who was camped near who had a satellite connection. Dragging the entire Love Machine over I was able to get the Python serial library downloaded, installed, and working.
With actual users on the Playa, I realized that the UI I designed was flawed in the sense that it does not alert the user that the names and emails have been accepted. This caused issues where people would tab through and then hit enter, which in Chrome causes the downloads to be shown. This caused some minor inconveniences for those running the booth the Love Machine was inside of, but nothing that a volunteer could not overcome. The other big issue I encountered was a double copy-paste error that resulted in about half of the data not being saved, where a function was named twice (like this, namedTwicenamedTwice).
As for the future, I don’t really have any plans for such a thing. I plan on doing a few more tweaks and as the Love Machine will be attending a few local regional burn events. The kiosk will hopefully be repurposed into something that can be used to better the regional burner community. This project was a lot of fun, especially being able to see the way people interacted with a machine that measured love. I will be doing another write up and project with the data that was collected during this project.
The code can be found here: github.com/evinr/Love-Staring-Machine