More selected projects

"Resuscitating Annie" 

Often neglected and dumped in cupboards, bags and boxes, a family of lifesaving dummies exists in our communities. Witnesses, spanning and defying time and metaphors for people who feel marginalised and unwanted. This project brings them to the party and celebrates their significance, importance and impact. As the analogue and digital balloons inflate and deflate in rhythmic harmony, the audience is invited to join the party through machine learning techniques and sensor driven interactions. 


produced by: Chris Newth

Introduction

Imagine the scene - It's 1981 and another party is under way at the Gasworks Social Club in Folkestone. However, the resuscitation dummies remain in the cupboard, out of sight and out of mind, despite their importance. Neglected. Marginalised. The same at the local scout huts, cadet units and schools where other dummies live; uninvited to the parties that take place in their homes. Punctuated by silliness, eeriness and placemaking nostalgia, this project uses computation to create a party for the dummies, inviting the audience to join in and for memories to be triggered and shared with the other party guests. 

Concept and background research

My project is in response to a placemaking initiative called 'Pioneering Places' and the former gasworks site in Folkestone, which is one of the Pioneering Places sites. I became particularly interested in the community around the site and the community hubs such as scout huts, cadet units, the college, and fire station. There was also a social club that remained on the derelict gasworks site, long after it closed. So, I knew I wanted to do a project about the gasworks site, but I wasn't sure of the exact details or concept at the start (May 2021). 

So, I began to share my thoughts and ideas with a local placemaking artist (Matt Rowe) and between us we came up with the idea of focusing on the civic effigies that are often found in the scout huts, cadet units, social clubs etc, but are dumped and discarded in cupboards.  For example, resuscitation dummies that the scouts and cadets might use for First Aid training or that might be in the social club for St John's Ambulance courses. Matt had already started a project collecting and photographing these effigies and offered to lend me a vintage dummy for my Goldsmiths project. 

In particular, I wanted to acknowledge and celebrate the importance and significance of these dummies; I saw them as a metaphor for people in the community that may feel dumped, neglected, decommissioned and marginalised. I wanted to 'bring the dummies to the parties'. I imagined and wanted to recreate a party scene from the 80s, when the social club was booming.  I also wanted to reference gas, air, and energy and this gave me the idea of using balloons, which fitted in with the party scene. I was keen to synchronise the inflation and deflation of the analogue balloons with digital balloons and spheres, blending and blurring old and new technology, acknowledging how they can work together. This fitted in nicely with the theme of our show. 

Here's a few sketches and images that inspired me and interested me. The shop window is Matt's studio in Folkestone, where I would test the installation and install it again after our show. The initial idea was to make a cast of the dummies but I chose to use the original ones instead. 

Technical

Artefact 1 - Dummies with Balloons:

To inflate and deflate the analogue balloons I used two 6V motors (inflate and deflate), controlling the air flow with a valve / solenoid. I used a valve as I wanted to be able to hold and release the pressure. All this was controlled by an ESP32, with associated electronic components (motor driver IC, switching transistor etc), mounted on a PCB that I designed and had printed.  I initially used and soldered the components on veriboard, but wanted something neater, more robust and that could be mass produced if I wanted to expand the project. For example, if I wanted to have 6 dummies in the future. 

I used an ESP32, as I wanted Wi-Fi capability, as I'd be sending OSC messages to an Open Frameworks sketch and a p5.js sketch to inflate and deflate virtual balloons.  I had a master ESP32 circuit that controlled the motors for Dummy A and sent the OSC messages to a slave ESP32 circuit (that controlled the motors for Dummy B), the OF sketch and the P5.js sketch.  I created a local Wi-Fi network with my own router, to avoid potential issues with the college network, which can be glitchy and unreliable. I set up reserved IP addresses on the router, so I could assign these to variables in the ESP32 code and not worry about devices getting a different IP every time the piece was turned off/on. 

I needed a way to ensure the balloons wouldn't over inflate (pop) or over deflate and I did a series of experiments with pressure sensors. This proved unreliable, mainly because the pressure being sensed was that of the pump and not the balloon, so it was constant. To solve the problem of over inflation etc, I used a code based timer in the ESP32 sketch using millis ( ) . I worked out that a cycle of 40 seconds inflation, followed by 30 seconds deflation, followed by a 30 second rest period worked well and I ran the balloons like this for 3 hours to test it. Having a shorter deflation time ensured there was a bit of pressure left in the balloon, so it never became over deflated (and potentially burn out the deflate pump). Venting it to atmosphere for 30 seconds afterwards, brought it back to a datum / rest before the next inflation. 

I used Open Frameworks to create a 3D sphere / wireframe that was synchronised with the real balloons, and I used p5.js with the facemesh API to insert a dummy's face over the audience’s face and give them a virtual balloon which was also synchronised with the real balloon. I used facemesh, as it could differentiate between types of face and give each person a different colour balloon. I used OF because of it's ability to efficiently render and manipulate 3D graphics.

I created an interactive mode, which could be controlled by physical switches or a touchOSC interface (on a phone). I decided to use touchOSC, as I wanted to avoid lots of wires and I thought it fitted in with narrative, where people at parties are often on their phones. 

Artefact 2 - CPR Dummy:

I wanted to create a CPR style interaction so mounted a Force Sensitive Resistor under the skin of the dummy. This controlled patterns on a p5.js sketch. The data was sent from an Arduino, along the serial port. As pressure was varied, patterns and effects were seen on the monitor, all relating to mental health.  I wanted a way to change scenes that was linked to the actions and interactions of CPR, so inserted a Light Dependant Resistor up the nose of the dummy. When the nose was squeezed (or face covered), this cut off the light, which triggered the next scene. I replaced the FSR with a proximity sensor, as the FSR was damaged at the end of the opening night. I used an Arduino Uno and the serial port, as it's known to be a reliable and simple method of data transfer.  There was no need to complicate things with OSC etc. 

Artefact 3 - Audio Player:

I wrote a Processing sketch, using the Minim library, to create an audio player where the audience could learn about the project. I used Processing, as it can handle basic audio player functions and could potentially be run on a Raspberry Pi (although I used a MacBook in the end). Also, with Processing, you can export the sketch as an application in 'present' mode and load that up on login. 

Weekly Blog:

I wrote a daily / weekly blog post, capturing all the technical and project development over the 4 months. Please access it here.

Future development

There's already interest in installing this in Folkestone as part of the Folkestone Triennial. To improve it, I would like to make the balloons inflate faster, maybe using 12V motors with a relay.  I also want to find a way of capturing the face recognition images (when the audience becomes a dummy) and uploading or printing this out, to create a record / portfolio of 'party guests'. I noticed during the show that this was a popular part of the piece, so it would be good to improve it.

I also need more visual references to the former gasworks site and Folkestone, perhaps some archive photos or a showreel, projected on the wall / TV.  

I would consider separating artefact 1 (balloon dummies) and artefact 2 (CPR dummy), making them separate 'pieces'. I take on board (from Viva) that perhaps too much is going on.  However, the CPR dummy did generate lots of conversations and interactions, so I feel it has a place somewhere. 

I was given very small, old monitors from the Computing Department, so had to stick with those for the Goldsmiths Show. I would like to experiment with larger monitors and perhaps project some of the images onto a wall. I think this would have more visual impact. 

 

Self evaluation

I observed that people enjoyed the dummies and the party scene. It made people smile and generated conversations and memories. It worked best when I was there as 'the party host' and invited the audience to sit next to me for a chat and to join the party. This created an intimate social experience, which was perhaps needed after lockdown and Covid! The piece became slightly performative, and this added value.

The CPR dummy was also popular, although the FSR broke on the opening night.  Replacing it with a proximity sensor perhaps worked even better, as the interaction was gentler and covid safe (no touch required). 

The audio player also worked well, although I should have had a larger sign inviting the audience to use / discover it.  

I think my project was ambitious in the fact that it linked in with a 'real' placemaking initiative and the work of a local artist. It also had educational aspirations and connotations, as I want to use the project as a teaching resource (and inspiration) for my own students. Technically it was very robust, running for the full 5 days without the balloons bursting and without electrical components overheating / failing.  Artistically, it needs a bit of refinement and perhaps stripping back to focus on one main element (the balloon dummies perhaps). As mentioned in 'Future Development', faster inflation rates and larger monitors or projectors might elevate the piece.  

It was incredibly tough doing this project whilst working full time and balancing family commitments, as well as being 2 hours from London.  There was a significant amount of planning involved, getting me and the equipment to and from the show, as well as additional financial outlays with hotels and travel.  To fit things in around work, I had to work on my project before work, after work, during weekends and over the summer holidays. So, I'm proud to have pulled it off and I acknowledge the improvements that can be made before rolling out again at the Folkestone Triennial.  

References

https://culturalplacemaking.com/

http://mattroweportfolio.co.uk/

Goldsmiths Summer Session 19/05/21 - Remote interactions for performances and installations - merging physical computing and web technologies, by Clemence Debaig

Open Frameworks Essentials (Chapter 5) by Denis Perevalov and Igor Tatarnikov

Code for loading 3D models into OF taken from https://openframeworks.cc/documentation/ofxAssimpModelLoader/ofxAssimpModelLoader

Code for facemesh taken from https://learn.ml5js.org/#/reference/facemesh

Code for p5.js OSC taken from https://github.com/genekogan/p5js-osc

https://github.com/genekogan/p5js-osc

Code for FSR adapted from https://learn.adafruit.com/force-sensitive-resistor-fsr/using-an-fsr

Code for proximity sensor taken from https://create.arduino.cc/projecthub/csbenk/range-finder-using-ultrasonic-sensor-563598?ref=user&ref_id=55932&offset=0

Code for serial control software taken from https://github.com/p5-serial/p5.serialcontrol/releases

Code for word / colour vectors taken from https://gist.github.com/aparrish/2f562e3737544cf29aaf1af30362f469

Code for p5 serial control taken from https://github.com/p5-serial/p5.serialcontrol/releases