Final Project: Family Dinner

• gwendolyn

FAMILY DINNER

Family Dinner is an interactive installation and a dark object. It is a two-person interaction that facilitates emotions of isolation and guilt. Each participant plays a specific role. First, there is the role of the elderly parent, who experiences having a dinner alone. While doing so, they hear a series of phone calls. Each time, a question is asked – “Are you coming home for dinner today?” — and each time, the parent is given a different excuse. Unbeknownst to the first participant, the second participant, who plays the child in this scenario, activates these responses by pressing various buttons on the opposite side. This experience is an interaction that drives contemplation about neglect and responsibility. Through the installation, we hope to provide a reflective encounter on parent-child relationships by eliciting intrinsically human emotions and desires.

DESIGN PROCESS

Initial Ideation

As mentioned in our previous post for project development (view here), our initial ideas were quite different from our final project. We initially wanted to highlight poverty faced by the elderly, and elicit the emotion of alienation they faced from society. Eventually we got stuck because we focused too much on the object rather than the emotion — we were trying to force an emotional encounter onto a set object when we should have been thinking about how the interaction can bring out an emotion.

 

After choosing to focus on isolation, we came up with different ways to facilitate this emotion (elaborated more in the project development post), but eventually we settled on having a two-person interaction based around a dinner table set-up.

 

Bodystorming

For a more detailed post on our bodystorming exercise and reflections: click here. The bodystorming was meant to help us test out our project, but in a really rough and simple way, just to get a gauge of the participant’s responses. We made our set-up out mostly out of cardboard and paper.

  

In theory, this is how the mechanisms would work for the set-up we used in the bodystorming exercise:

We gained a lot of valuable insights and feedback from our classmates during the bodystorming. Here are some main takeaways from the bodystorming:

  • The child-participant was initially just meant to sit and watch the other participant eat alone – not actively partaking in the interaction.
  • Additional set of bowl and chopsticks may be confusing — participants do not know where to sit.
  • No clear indication of the end of the interaction, participants did not know when to leave.
  • There needs to be a clear set-up, with deliberate props to help immerse participants in the context, so they can make the conceptual leap to the message instantly.

Refinement & Development

After bodystorming, we refined the installation in several ways:

  1. Additional roles for the participant playing the child – trigger responses to the recording
  2. Sourcing for props that could would help participants make the association to the elderly
  3. Refinement of voice recording to indicate conclusion through dial tone

Coding Process

Parent’s Side

After consulting Lei & Serena, we learnt that we had to use Processing to trigger the recording to play from our laptops. I did some research on how to get Processing to play sound.

For the Arduino’s side of code it was pretty simple, we just had to get the Arduino to print the values from the photocell, and if it was more than a certain threshold, it would print a speicfic letter. This would trigger Processing to play the sound.

The first Processing code I found for playing sound: https://textuploader.com/1d5md

This was quite straightforward, all I had to do was replace the name of the audio file and it was good to go. The only thing that I realised I had to amend was to remove the serial print command for the actual photocell value because it was causing pauses when Processing played the file.

Child’s Side

To create a stronger sense of narrative, and to highlight that the isolation the elderly feels is a prolonged one that exists throughout their child’s life, we decided to have three responses. Each one corresponds to a certain phase in the child’s life, from being a student, to going to the army, and then finally becoming a working adult.

Our initial idea was to put a photoresistor under each shirt, and when the participant lifts it up, it triggers Processing to play a specified recording. However, we thought it would be quite counter-intuitive for participants to have to lift the objects up, especially since this was an installation and they might be hesitant to touch anything laid out properly on the table.

We also thought of having a phone that sends the responses when the participant taps the button, but the association to the different phases of life would be less strong.

In the end, we chose to stick with the shirt idea, but change it to three buttons instead. We thought buttons were more intuitive — the participant would immediately know how to interact with a button.

For this interaction we required three different inputs with three different outputs. So, we couldn’t reuse the code from the parent-participant’s side. I did a little more research on how we could get Processing to play specific recordings with specific inputs.

I found out that we had to add this chunk of code at the start so that we could tell Processing to play different tracks later on in the code.

For Arduino, it was the same as the parent-participant code, except now there were 3 inputs that would print 3 different messages in the serial monitor when triggered. Then I moved on to the circuitry:

This was the first circuit that I set up when our idea was still to use the photoresistor. I thought I could just replicate it and replace the photoresistor with the buttons, but it didn’t work because the buttons weren’t just sensors, they actually opened and closed the circuit.

 

I had to redo the circuit, following one that Serena found – this one worked with the new code:

 

Play Test

The play test helped us to check the responsiveness of the various sensors. The circuits worked! The main takeaway we got from this was to change the interval between each phone call from 15s to 10s, since with the play test we got to listen to the entire recording and observe the reaction time from the participants, and we felt that 15s was a little too long.

Also, it was after the play test that we got a suggestion to adjust the lighting (bright to dark on parent’s side and vice versa for child’s side) to enhance the finality of the interaction’s conclusion.

CREATION OF FINAL INSTALLATION

Step 1: Turning a piece of acrylic into a one-way mirror

Apply the one-way mirror film on the wrong side, realise it’s the wrong side and re-do it (oops). Using a flat edge, like a scraper, smooth out any bubbles. Then, using a penknife, poke holes in remaining bubbles trapped in the film to release the air.

Step 2: Prepping the table for the acrylic

Mark and pre-drill holes the size of the screw into the table (this took us awhile as the wood was really hard to drill into). Secure the L-shaped brackets with screws. Slide the acrylic in.

Step 3: Drilling holes for the sensors

Mark and drill holes for the photosensor and three buttons.

 

Step 4: Set up props

Cut a slit down the middle of the tablecloth and slide it in through the acrylic. Also, cut holes in the tablecloth for the sensors. Try out the arrangement of the cutlery to see which feels the most natural. Lay out the three different shirts.

Step 5: Set up circuitry

String sensors through the holes, tape the Arduino and breadboard onto the leg of the table. Tape the wires up to the underside of the table to conceal them. Tape the USB cables onto the leg of the table and down to the ground.

  

Step 6: Set up LEDs

Tape the LEDs onto the edge where the tablecloth meets the acrylic. Tape down the wires securely on the ground to ensure the dimmer stays in place.

 

Step 7: Print out instructions

Attach the pieces of paper onto the table and buttons with clear tape.

 

Step 8: Set up backdrop

Set up the backdrop so each participant can enter their respective sides without seeing what’s on the other.

   

USER TESTING

In total, we carried out three user tests before the actual presentation. We tried to get people who were not taking this class, or had not heard/tested our project before to ensure that we got the most honest and genuine reactions. This was also a good way to see if our set-up was intuitive enough for participants to understand its mechanisms, and immersive enough for them to make the jump from context to message.

For the first test, we did not include any instructions on the table. We observed that it was not instantly clear what each participant had to do, so we gave them verbal instructions. For the participant playing the child, we told her specifically to press the buttons in an order from left to right. The sequence was particularly important for our project, as it showcases the passing of time and creates a sense of narrative, from when the child was young to when he is grown.

TEST 1 – Parent’s Side

TEST 1 – Child’s Side

After this first test, we decided to include simple instructions. First, it was for the parent’s side, to tell them to pick up the spoon to start eating, and then for the child’s side to guide them along the sequence of responses.

TEST 2 – Parent’s Side

TEST 2 – Child’s Side

TEST 3 – Parent’s Side

TEST 3 – Child’s Side

After adding the instructions, both participants could clearly understand how to interact with the piece, without us telling them anything at the start. Here are some responses given by the testers:

  • In general, they all felt that the switching of lights at the end was a clear indication that the interaction had concluded.
  • Previously, we had allowed for 15s between each call, but cut it down to 10s. The user test was a good way to check if the duration was too long or too short.
  • We were also worried that the person playing the parent would not be occupied enough as compared to the child, but we received feedback that they were focused on listening to the conversation and waiting for the response, so it wasn’t boring even though they didn’t have much to do.
  • The participants who played the role of the child also said that the combination of the button pressing and the fact that they were able to see the other participant amplified the sense of guilt.

REFLECTION

I learnt so much through this final project. This whole process of creating an interactive installation has taught me valuable technical skills, such as coding and circuitry, but more importantly I have learnt that it takes so much to curate an effective interaction. From the choice of props, to the placement of objects, to sound and environment, there is so much we need to consider when it comes to creating an encounter. Every little detail is crucial to create an immersive experience — as designers we play the role of guiding the participant to discover the message through interactivity. This project has helped me to understand both the keystones and nuances in making an impactful interaction, and I’m really happy with the final installation 🙂

RESOURCES

Codes

Audio

Leave a Reply