As described by designers Severino Alfonoso and Loukia Tsafoulia, "Synesthesia is an atmospheric, soft interface with an infrastructural core that allows no physical bodies inside but rather consumes them virtually. It is a manifestation of bodily data relationships abstracted and projected back to an analog domain.".

​

This project was conceptualized at the Synesthetic Research & Design Lab at Jefferson University. I collaborated with the lab leads, Severino and Loukia, to bring the technological aspect of their creative vision to life. I was presented with a design brief and feature request from which I designed the technical architecture and bill of materials.

​

The 10x10x10 ft. body houses 4 Raspberry PI computers networked together using a Pub/Sub protocol model. The system is designed using distributed computing principles, which proved to be extremely efficient and performant for the installation. 

​

There are 3 different services powering the application:

1. Proximity detection and video capture of the eye (Interaction Service): detects when there is a person looking through the surface and turns on the video camera to record the person's eye.​​

2. Light control (Light Service): receives proximity event from the Interaction Service(s) and changes light patterns and intensity to transfer to a more agitated state.

3. Sound control (Audio Service).: received proximity event from the Interaction Service(s) and changes sound volume and intensity to transfer to a more agitated state.

​

The cameras and computing units were physically placed close to the surface of the Synesthesia body, mounted in 3D printed modules that were designed and assembled in-house. 

​

​

Challenges of building an installation that travels the world

​

Synesthesia was designed to be presented at the Venice Biennial 2021 and travel around the world. This came with interesting challenges. For example, the installation needed to automatically start all its systems, connect to the network, and be fully functional whenever it was connected to power. To achieve this, I modified the Linux startup scripts to hide unnecessary UI elements and encapsulated all 3 Python applications as a Systemd service. To remotely debug the installation in case of failure, all devices are connected to Dataplicity, a remote SSH reverse-proxy system.

​

Heat management was also a key aspect to get right. I sourced the best hardware power supplies and components, managed their power consumption, and provided enough ventilation to ensure that all systems could run for a long period of time.