Category Archives: Physical Computing

[Thesis] We’re Still Here


My thesis is entitled We’re Still Here, and it is the beginning of a planned series of ordinarily mundane objects that have evolved personalities and behaviors that are based on their original functions. Here’s the longer blurb:

“Why should our sleek, sexy, status-symbol gadgets get all our attention?

We’re Still Here is an exploration of the ordinary objects in our lives that perform their duties day in and day out without much acknowledgement or conscious thought from their users. Each object in this collection is modified to display surprising behaviors or personality traits that are derived from how it normally operates; the series begins with a neurotic, overly needy alarm clock and a dutiful-yet-exhausted coatrack that just wants to catch a short break.

By giving personalities to these objects, I’ll playfully invoke a new way to look at and think about the myriad commonplace, “boring” tools that quietly contribute to our lives.”

I’ve been interested in the idea of personifying objects since my beginnings at ITP; my physical computing final, The Embarrassed Book, was based on the idea that a book may not always be necessarily interested in divulging the information it contains. It also fascinates me how, out of the large variety of objects we use everyday, a select few are extremely predisposed to become status symbols while the vast majority have no chance to ever really capture our attention or imaginations. I want to help our largely unconsidered tools become more interesting to us.


The Alarm Clock:
The alarm clock consists of the shell of a real alarm clock fitted with a seven-segment LED display, two piezobuzzers, and capacitive sensing controlled by an Arduino Pro Micro. It will be extremely needy, desiring constant touch, and will count down to a specific time at which it will cry out for attention and express its disappointment with you. If you happen to touch it independently of its time limit, it will chirp pleasantly and display its satisfaction.

The concept for its behavior is based on the idea that a normal alarm clock is essentially already a highly neurotic object that goads you into touching it regularly at a very specific time every day. Users depend on alarm clocks to wake them up, but they never consider what the clock’s needs might be in this relationship.

Pictures of construction:

Video of an early prototype in action – no sounds yet, but displaying positive messages:

The Coatrack:
The coatrack is interested in performing its duties as a coatrack well, but gets physically tired from holding up bulky coats all day. Thus, when it thinks you’re not around to watch it, it likes to slouch over to take a small break. The concept is based on the idea that a coatrack stands perfectly straight and bears a heavy burden for its entire existence, performing its duties admirably, and yet nobody ever really notices or appreciates that it’s doing so. It must be a thankless lifestyle.

The coatrack will consist of segments of PVC pipe with 3-D printed caps at the end that will allow 3 cords to run through them. At the base of the rack are three high-torque servos that will hold the cords taut for a straight appearance, or selectively allow slack in different areas to let the segments lean in a specific direction. Photocells under the three hooks will let the Arduino know when coats are hanging and decide which servo to let slack. The rack will detect if people are approaching using two IR rangefinders.  Some concept/testing pictures:

And a video of the most recent prototype with the full-size PVC segments:

Development & Documentation

The Alarm Clock:
Since the prototyping phase, I’ve refined the clock’s behavior to make it more believable and engaging as an object. Beforehand, it started out with a countdown to when it wanted to be touched; now, it acts as a normal clock for a random amount of time, and it switches to the countdown when it starts to get impatient with your lack of attention. These behaviors are both present in the documentation video, just shortened to keep the video at a reasonable length.

You can see the basic behavior from the video, but not the variety of messages it displays. Ignoring it will always trigger a regular alarm clock beep, but the clock will choose from a pool of guilt-tripping messages including “don’t leave”, “deserter”, and “you’re cruel”. Appeasing it will trigger one of a variety of happy chirps and messages such as “again, again” and “ahhh, so good”.

The Coatrack:
The coatrack is still a work-in-progress, as it is more physical/mechanical in nature and thus a harder problem to solve. The idea I had from the prototyping phase was that when it sees you, it stands up straight, but when it thinks you’re not around, it slowly starts to slouch so that it can relax a little bit. I have encountered some mechanical issues that have slowed progress – namely, putting even a light load on the hooks inhibits its ability to stand up straight.  You can see a short clip below with a sampling of its different movements both with and without a coat on its hooks. I hope to continue working on it over this summer to improve the design and advance it to its original planned state.

The Presentation

Here, you can witness a view of my slides and listen to my discussion regarding my thesis that I presented for ITP Thesis Week. It’s roughly 14 minutes long and includes both the video segments shown above.

Pop-Up Window Display – Sock Monster


For our 7-week Pop-Up Windows class, we were tasked to form teams and design an interactive window display for a vacant building on the NYU campus. During this time, my team and I planned and installed a window exhibit in which a sock is lowered into a giant washing machine and is eaten by the monster that lives there. The project involved animation, puppetry, fabrication, set design, projection, and physical computing. We built the washing machine by stretching fabric over a wooden frame with an open back so that we could project onto its front. A button was placed on the outside of the window display, and when it was pushed, a stepper motor lowered the sock into the machine, and then animations played of the monster interacting with, and ultimately pouncing upon, the shadow of the sock. After that, the lights would dim on the display and the sock would retract to its original position, ready for the next user.



The Sock Monster display next to its neighbor The Internet.


An Arduino controlled the stepper, the DMX lighting, and the button control, and interfaced with MAX/MSP to randomize and play the different animations at the appropriate times. My role on the project was primarily creating the monster’s animations and doing the MAX/MSP programming. I also assisted with the construction of the frame and with the Arduino code as well.

Here’s a timelapse video showing an early stage in the construction process – our team comes in in the second half of the video to start work on the frame:


This Drawer

Our midterm project for Poetry Everywhere was very open-ended: the assignment was to interpret a poem into a new media format.

Xuedi and I teamed up once again. We chose to make a project in response to Ron Padgett’s poem “Nothing In That Drawer”  because we liked the extreme use of repetition and the mixed feelings of amusement and exasperation that comes along with that repetition. Our initial project ended up splitting into 2 separate pieces, and I’ll talk about them separately below.


This Drawer:

We painted and wired up a 26-compartment storage container so that we could detect when each drawer was opened by a user. We did this by putting copper tape on the back of the bottom of each drawer, and the front of its housing. When a drawer was open, the copper strips would touch, and they would be separated when the drawer is closed. By soldering wires to each of these copper strips, we were able to connect them all to an Arduino Mega and use them as simple switches to detect whether any individual drawer was open or closed.

In our first iteration, we put nothing in most of the drawers, and put stanzas of a response poem that we created (detailed further below) in the remaining drawers. All drawers with nothing in it were set to produce a different audio cue of me saying “nothing in that drawer!” every time they were opened. I tried to make my voice gradually progress from disengaged to distressed over the overall course of the different cues.

We observed people’s reactions to the drawer at the ITP Show, standing nearby but far enough that people wouldn’t assume we were associated with the project. The vast majority of people seemed amused, perplexed, and impatient, which mirrored some of the emotional reactions we had to the poem itself.


In the first iteration, we decided to have 14 empty drawers (a reference to the number of lines in the poem) and 12 drawers filled with different stanzas from our response poem. The class feedback we got seemed to indicate that we should separate the drawers and the poem into two separate pieces. In the second iteration (shown above), all 26 drawers are empty and we made more audio cues to reflect that fact.


Upon Finding Nothing In The Drawer:

This is the response poem that was originally found in the drawers in our first iteration. Our intention for this part of the project was to inflict nothingness upon people, see how they would respond, and compile those responses into a poem. We achieved this by bombarding people we knew (and in a few cases, total strangers) with blank texts and blank emails. We also created a mostly blank website and asked the ITP community for feedback on it. We then curated and grouped the responses we got by general sentiment and created a poem out of them.

The poem is as follows:


Upon Finding Nothing In The Drawer
Compiled by Andrew Cerrito and Xuedi Chen


This is sort of annoying, really.
What are you, a fucking ghost or something?
Fuck, it’s empty. Bastards.
I hope you’re laughing right now.


I’m kind of seeing nothing. Is this art?
You gave me nothing. Is this conceptual?


Are you doing secret lemon-and-a-candle messages?
Lorem ipsum dolor sit amet.
I just see bubbles.
Yeah, don’t remind me.


What’s going on? Is everything okay?
Are you hiding something?
What do you want? What’s the big secret?
What are you trying to say?


I looked around more,
but I was still disappointed.
Nothing is working.
This is tearing me apart inside.


I feel partly pleased and partly confused.
But I swear I saw something.
I kind of like that.


(no response to nothing)
(no response to nothing)
(no response to nothing)
(no response to nothing)

ITP @ Maker Faire – Ohm Wrestling

I was a part of ITP’s team to develop a large-scale project for Maker Faire 2013 in New York. A description from project manager Hannah Mishin: “Ohm Wrestling is a human powered, collaborative mechanical arm wrestling competition.

Two mechanical arms are setup in arm wrestling position for teams Blue and Orange. Each team has an array of energy harvesting devices behind them feeding into the corresponding arm (their motions being Shake, Push, Pedal, Pull, and Crank). The various devices show different methods of harvesting kinetic energy from human motion, energy that the participants use to power their team’s arm. The harder the teams work, the more power they produce – giving more force to their arm to win the competition. In addition, each device is equipped with a light to show each user their own contribution to their team’s winning the competition.

Variables that contribute to a team winning the competition include the amount of watts produced as well as the endurance of the team (how long they can last at peak energy production levels). A large meter reflects the teams’ status/energy produced for the competitors and the audience.”

A photo of the entire exhibit at Maker Faire. Photo by Natasha Dzurny.

A The entire exhibit assembled at Maker Faire. Photo by Natasha Dzurny.


This was an extremely intensive project that took a lot of time and energy from a large team of people. My role in the project was to create two machines with fairly high energy-generating capabilities. I decided to make a pair of large, 4′ x 4′ elevated platforms with two aluminum poles in a T-shaped configuration coming out of the middle. Users could grab the horizontal pipe as handles and push the handles in a circular motion around the platform. Underneath the floor was a very large wooden gear connected by scooter chain to a 40W motor with a small gearhead. The resulting large gear ratio allowed users to generate a fairly large amount of electricity compared to the other devices on display.

This was my first large-scale fabrication project, and most of it was non-digital to boot! It certainly was a challenge to work with gearing, mounting the pole mechanism securely, and creating a durable object that could stand up to the abuse of a great many children. Unfortunately, my computer and camera with most of my documentation of this build were stolen from me before I could back it up, but some pictures of the platforms being constructed/in use are below:

And, finally, here’s a shot of the awards that our exhibit won:


Exhibit photo taken by Natasha Dzurny.

Platform photos taken by Natasha Dzurny, Talya Stein, and Caroline Sinders.

Awards photo taken by Caroline Sinders.

Peanut Gallery

Peanut Gallery was my Live Image Processing and Performance project that also unintentionally became my Digital Fabrication final.

I knew that, for this final, I wanted to do something physical with video; I didn’t want to make a project that you simply observe and don’t actively participate in. At the time, I also was interested in trying to create my own inflatable figures, and figured that would serve well as an interesting video interface.

From these ideas, I came up with the concept for Peanut Gallery, which is an interactive TV movie watching experience in which the occasional obnoxious person pops up next to you on an inflatable tube to disrupt your concentration.  The different people that pop up can adjust your TV’s signal in different ways according to their own preferences. Some of these disruptions include adjusting the colors or the contrast, rewinding the movie, muting the movie, changing the channel, turning off the TV, or turning down the volume. To restore peace and order to your TV time, you must strike the inflatable with the interloper on it until it deflates.

An example of one of the disruptors – he turns off the television completely.

I started by fabricating the enclosure out of plywood on the CNC and sewing cylindrical shapes out of ripstop nylon. I hooked up IR rangefinder sensors inside each inflatable near the base so that I could sense when a figure was hit by a user, and I inflated each of the 2 cylinders with a reasonably powerful 12V boat fan. These were hooked up to an Arduino that interfaced with a Max/MSP patch that controlled both the projections mapped on the inflatable tubes and the video effects applied to the movie on a nearby television. I used a dual head video splitter to route video to both the television and the projector.

Unfortunately, not much media remains of this project because the computer I was storing it on was stolen (a painful lesson in backing up my content more frequently!), but some pictures and video remain. Here are two photos from the ITP spring show:

And here are some (very short) videos of people interacting with it:

My friend Alex really going at it.


These kids stayed to play for about 5 minutes – probably my favorite moment over the project’s lifespan.

The Collective DJ – Winter Show 2012


The Collective DJ

Created by Andrew Cerrito, Mary Fe,

Colin Narver, and Azure Qian



The Collective DJ is a project that allows people to create music through touch and collaboration.

Four different sets of two handpads each are mounted on a wall, spaced apart from each other. When two users each put one hand on a handpad and use their other hand to touch one another, they can trigger an audio sample to turn on or off. Since there are four sets of handpads, up to four samples can be triggered at once, allowing multiple groups of people to DJ a track together.

How it works:

When two people connected to the pads touch each other, a small amount of electricity rushes through them and they complete a circuit. The circuit completion is detected by a microcontroller, which uses keyboard emulation to send one of four keystrokes to a computer running Ableton Live. In Ableton, the four keys are assigned to different samples of an audio track, and those samples are turned on or off via keypress.

Iteration history and design considerations:

This project was actually the second version of The Collective DJ. The first version was a project created in ITP’s Applications class as a reaction to a talk given by Nathan Shedroff on sci-fi interfaces. The talk got us thinking about what kind of novel interfaces we could design, and since our presentation was in the middle of midterm projects period, we wanted to give our class a fun and positive experience.  This version was designed for a large auditorium, and it involved 8 rows of floorpads that triggered 8 different sounds and effects. It was used by the entire 2014 class of ITP, which is approximately 110 people. (A video is posted below.)

For the Winter Show version, we had a much smaller space to work with – a narrow hallway in the back section of the ITP floor. We decided not to use floorpads, as people would have to take off their shoes – an action that would both dissuade some people from participating and create a traffic jam in the busy hallway. Wall-mounted hand-pads were much more appropriate for the audience and the space.

The first iteration had 8 sets of floorpads and 8 samples, but we scaled down to 4 in order to make the installation easier to comprehend at a glance and to make sure each sample could be distinctly heard amidst the crowded, noisy atmosphere of the show.

Our wall space had both a long unsightly whiteboard that couldn’t be removed and a kiosk window. In order to disguise these, we papered over the whiteboard with heavy black paper and nailed a large piece of dense white matte board over the kiosk. For visual contrast, we mounted the white hand-pads on the black backing and vice versa. The pads also had a lasercut hand design in order to provide visual context to the users. We mounted the wires to the wall using nails and tape so that they were organized and not visually distracting (a big problem with our first iteration), and we directed them to our computer station to the side of the installation.







Video of the first iteration:

The Embarrassed Book


The Embarrassed Book is a shy book that is reluctant to divulge its contents. It will sense when a user opens it via a flex sensor in the spine, and, as the user reads, it will give mild audio cues cluing the user into its discomfort. Eventually, it will get too embarrassed, shriek “Don’t look at me!”, and snap its own cover shut via a motor embedded in the back half of the book. It is important that the book resembles a normal book so that the user will not suspect something upon their first encounter. Ideally, the book will be a self-help book about asserting yourself, getting over shyness, or something of that nature.


I started prototyping by hollowing out the pages of the back half of a book to form an electronics compartment. Cutting by hand with a boxcutter was time-consuming and difficult. Pictures are below:

The yellow wire (a prototyping substitute for clear fishing line) was attached to the inside cover, and was threaded through the spine of the book into the secret compartment at the book’s end. Unfortunately, when I pulled the wire, the book cover would not close. This is because I needed some upward force to the pull, but because the compartment was by necessity lower than the inside cover, this seemed difficult to achieve. After spending quite awhile trying to think of a different way to close the book, I was unable to come up with one that wouldn’t unduly compromise the book’s normal appearance.

Iterating and Building

After considering my options, I decided to construct a lectern for the book that will house all the electronic components. This is different from my original vision of a self-contained book, but at least this way the book can remain unaltered and have a completely normal appearance. There will be a slot in the lectern through which a tiny arm can come up and push the book cover closed at the appropriate moment. The arm will be powered by a high-torque servomotor. Instead of using a flex sensor in the cover, I will now use a photocell embedded in the lectern to detect if the cover has been opened.

After constructing the base, I had to make the top plate of the lectern. I cut a slit in the top cover of plywood using the chopsaw, and made it fairly wide so that the servo’s plastic mount and arm would fit through it. I also drilled a small hole in which to embed a photocell, but made it smaller than the cell so that it would not sink through. I then lasercut a small pushing arm (about 5″ long) out of acrylic to attach to the servo, and mounted the servo to the underside of the plywood so that the arm could push upwards through the slot.

After everything was in place on the lectern top, it was functional but very ugly – the slit cut roughly, there were a few errant screwholes, and the plywood itself wasn’t in the best shape. To attempt to make it look a little nicer, I bought a large piece of self-adhesive foamboard, lasercut the slit and photocell holes in it, and attached it to the top of the plywood. Unfortunately, to fit in the laser cutter, I had to saw the foamboard in two which created fringing along the edges. In the end, though, it still looked marginally better than the plywood. I finished construction by affixing some leftover plywood strips to the lectern to serve as a book stand. Here’s a picture of it as it was presented in class:

The Embarrassed Book

Wiring and Code

The wiring for the book’s circuits was relatively simple, as there weren’t that many different elements in play. Since the servo required its own power supply with higher amperage than the Arduino could handle, I split the breadboard halves into a servo half and a photocell/mp3 trigger half. Here’s the diagram in Fritzing:

Book - Circuit

The program to control the book used a boolean logic system and timers to keep track of when the book’s cover was opened, when audio cues should play, and when the servo should close the book. The code is on Github and can be seen here.

Final Documentation

Build Issues

The biggest issues I had in this project were mainly related to the mp3 trigger. When I had first soldered header pins to it and hooked it up, I noticed that it would not maintain a constant power connection – jostling it even slightly would cause it to turn off. I figured that the header pins I had used must have been loose, so I attempted to desolder them. In the process, I managed to scorch the VCC pin and the trace leading to it.

After that, I had to hook up the trigger to external power, but I could not get it to work properly no matter what I did. I thought it might be a software serial issue, so I checked and rechecked my programming, but it wouldn’t initiate communication even with the simple example provided with the Arduino IDE. I thought that the trigger might be shorting out somewhere due to the damage I gave it, but I tested all the relevant pins with a multimeter and they all appeared to function normally. In the end, after a lot of stress and wasted time, it turned out that the firmware was simply not up to date. Lesson learned – try the simple solutions first!

Other than mp3 trigger troubles, I didn’t really bump into any major problems during construction and testing. I did manage to ruin an Arduino screwshield by following the wrong instructions online, but I’m not going to elaborate on that because it’s embarrassing.


Once I shifted focus from the electronics being entirely self-contained in the book to using a lectern, the project became much more basic – it’s just a photocell, a mounted motor, and an mp3 player, and computer speakers. Nevertheless, I did learn some important skills along the way. Laser cutting in particular was a pleasant surprise – it’s relatively simple to setup and immensely satisfying to execute. Software serial emulation seemed like a confusing concept at the time but ended up also being pretty simple, and it seems like a powerful tool for future projects. And after leaving a wake of destruction, my soldering skills have reached an acceptable level, and I got to use a few new shop tools to boot.

If I expand upon this idea later, it’d be neat to have different books have different personalities and reactions. Although I used a Processing instructional book for prototyping, my intent was to have the embarrassed book be a self-help book about learning to assert yourself. It’d be fun to have some other books around as well – maybe a prudish art book that censors itself once you reach a nude portrait, or an old antique book that suffers from senility. This could be done using RFID tags and a reader.