Reflection and Documentation of work from DECO3850, Physical Computing and Interaction Design Studio course by Dimitri Filippakis


My name is Dimitri Filippakis and I am a undergaduate student at the University of Queensland studying Bachelors of I.T. majoring in User Experience and Information Systems. I have a wide variety of skills within the industry, which is very handy when working in projects. With studying I also enjoy getting hands-on with labour work and fixing things around the house.

“I Strive for two things in design: simplicity and clarity. Great design is born of those two things.”


The Ideal

The domain space that team CDI and I have been focusing on this semester is enhancing mundane spaces. This domain is about dull, boring spaces that lacks excitement or enjoyment, for example these can be buses, elevators and waiting rooms. Inside these spaces you most likely find people bored, aimless scrolling on their phones. With this in mind, the teams focus became to help eradicate this boring stigma surrounded by elevators and the space that elevators are within. In doing so we aimed to hopefully get people off their phones and get them moving, be happier, or to just simply entertain themselves through a physical and digital interaction. My project explicitly seeks to follow these guidlines of making the elevator a more exciting space, whilst also being focused specifically on residents of aparatments. In doing so I want people to move beyond their comfort zone and make their everyday interaction with an elevator to become more thrilling.

Specifically, we produced two ideas for a physical interaction, one for a single user and one for a group of people. The single user’s ideal interaction starts after they select their floor, before the elevator starts moving the elevator gives the user an image with someone posing on screen in which that would have to mimic. If they can copy the pose then the elevator continues to the floor they selected, if not than they have unlimited attempts to try and match the pose. The ideal experience for multiple users is that once more than one person enters the elevator, the elevator detects this and presents the users with an audio output (from built in speakers) explaining the situation. Once that is completed, it gives one user a specific word (which is displayed on screen). With that word in mind they must act out for the others the word so they can guess what the word was (just like a charade game). Once one user correctly guesses the charade word out loud, the elevator hears this and lets the users select desired floors.

Below are illustrations of the ideal oncept. On the left we have the single-user interaction and the right is the mulit-user interaction.

Single User Interaction

Multi-User Interaction

The Actual

Single User Example

Multi User Example

Below are two links to the working prototype, the first is the single user interaction and the second is the multiuser interaction. Below we have an image of the final build (the elevator) and to the right we have both interfaces that users see on the screen when they are within the elevator. Alongside those is a quick video showing the prototype being used as well as a quick explanation of the product.

Single User Demo
Multi User Demo

“Two of the most important characteristics of good design are discoverability and understanding.”


The Build

The prototype is made with two parts combined. There is the physical build and the digital build.

Physical Build

First there was the physical build. This incorporated making a squat-rack out of wood than using sheets to enclose the sides to make the inside feel like a small rectangular prism (just like an elevator). This was done following Buff Dudes Tutorial on YouTube. The next step was displaying the computer screen into the elevator, this was done by mimicking what was shown on the website onto a spare iPad. By then installing the iPad onto the squat rack and using Splashtop the website would be able to duplicate into the ‘elevator’ for the users to see. The next step was getting the button to work like a real elevator. By using an Arduino multi-button welded onto the board and connected to the breadboard connected to the Arduino Uno, when a user clicks on their ‘level’ it triggers it so that an interaction would load. By clicking quickly, it opens the single user prototype and by holding the button it opens the mulit-user prototype.The Arduino is than connected to the computer alongside the webcam. The webcam acts as a camera to read the users posing and as a microphone for the multi-user charades game. In creating a physical elevator, I wanted to be able to simulate a real elevator experience (due to COVID-19). In doing so I wanted users to feel and experience how the product would be used in a real-life experience. I believe that in doing so I was able to get more of an emotional response from users when testing.

Digital Build

The digital build incorporates using HTML, CSS, JavaScript, Arduino Code, and Unity alongside Xamp and Splashtop. Once the button is pressed, the Arduino code either prints a 1 or a 2 to the computers serial port. Unity than uses this serial port answer to choose which website to open. The two prototypes for the different users are set up on websites on the zone. The single user uses PoseNet alongside a Cosine-similarity algorithm that helps detect the differences between an image and the live webcam. The Multi-user website uses a google speech recognition API and uses that to detect what is being said then with that, compares it to a list of randomly generated words. As stated above, Splashtop is than used to mimic what’s on the computer screen so users are able to see what they need to do. The image on the right is an image map of how the code works.

“Design is so simple. That’s why it’s so complicated.”
- Paul Rand


User Input

Throughout the development of the project, there have been several iterations which have changed the experience. The several prototypes have changed varying on either peer or user feedback. The first iteration included changing the original idea of ‘dance to move the elevator’ as it was pointed out that there would be a limitation of both space and time for users. This than pushed our project to be less physically active but rely more on user’s ability to think immediately. With this iteration we settled as a team that there would be two interactions one for the single user and one for multiple people. This brings us to the current concept of the pose matching and the charade like game for the elevator. From here, I conducted interviews with apartment owners in which they gave valuable feedback that has moulded the current prototype. This being, that using the elevator everyday becomes tedious when they want to leave and get home quickly. Alongside this, they made it extremely clear that whatever activity they would participate in would require them to be hands free.

The 1st iteration of the new concept guaranteed that the poses being conducted were able to be completed without the need for hand actions and that the charade game be easy enough that they can quickly participate and get to their floor within a minute(this would also depend on users peformance in the charade game). After testing for that prototype was completed, the results indicated that users did not like the need for touching/reading the screen and hence the current prototype was made in which the system became automated in which it both informs the user what’s required while also not needing the user to interact with the screen to continue. Without constant feedback from peers and users, the prototype would be completely different and would be lacking several features that is crucial for enhancing a mundane space.

Project Outcomes

At the products current state, it is in the ideal form to be installed within actual elevators. Although testing has been done with the current product, there is a big difference between a simulated experience and the real thing. Although this prototype was created to accurately simulate and elevator experience (due to COVID-19), there are still iterations to be completed. In doing so revision for the audio feedback and speech recognition would also be necessary to accurately inform and record users as there have been unknown errors in the past. In terms of meeting the ideal concept, the product does do so. It can check poses, and able to detect speech recognition for the charade game.

The user experience has been created in respect to the previously established domain. The domain being enhancing a mundane space, we aimed on creating a fun and exciting space for users to increase their enjoyment in a dull or boring space. Mimic Me focuses specifically on residents of apartments and getting a positive emotional response after the elevator experience. Hence, its evidently clear that Mimic Me reflects on the team’s domain.

Additionally, Mimic Me can be connected to the studios theme of “Designing for playful and open-ended interactions in everyday life: human values in physical computing”. The interaction being that users are needed to either pose or participate in the charade game to reach their desired floor. Both activities be physically demanding, requesting that users move to pose or use their initial thinking and surrounding area to complete the charade like game.

Mimic Me in its current state has been able to outline a set of actions that is required from users, hence able to easily communicate it is intended experience. In saying so, with the current audio clips, all users have been able to comprehend what is required of them for the elevator to move. All users were able to match the poses given to them and able to complete the charade game with the randomly given word. Overall, the final product has finalised into a very successful outcome when comparing it to the ideal concept that was created earlier in the semester, further modifications would need to be conducted to have a more accurate emotional response from users when the product is placed in their own residential buildings.

“The function of design is letting design function.”
- Micha Commeren


Throughout the course, to build the product we conducted, research in the domain, created a proposal, presentation and more. Links from previous individual and team work for this subject can be seen below

Team CDI

Alistair Harris
Anson Cheung
Dimitri Filippakis
Seamus Nash

Previous Work

Original Idea Video
Design Proposal Report
Prototype Video
Prototype Report