Category Archives: FinalProject

Use this tag only for your final project.

Spatial Musical Instrument for the Hearing Impaired


The aim of this project is not to ‘prescribe’ hearing impaired people a spatial instrument so that they can start interacting with music. There are already many cures to deafness and the currently incurable kind (sensorineural hearing loss) is not to be treated with the techniques being used in this project as it is a rather a neurological deficiency than a mechanical one. This project rather aims to create a setting where hearing impaired people and people with optimal hearing can generate and listen to music together without having the impaired individuals feeling like they need an “aid” or “fix” to fully be a part of a musical context. This is eventually an attempt to revert the marginalizing effects of biological obstacles on communally shared spaces and activities

Sci-fi inspiration: J G Ballard, Sound Sweep
My departure point was J G Ballard’s science fiction story Soundsweep where, due to the noise pollution in a distant future, contemporary musicians start shifting the sound range within which they compose into an ultrasonic level. Therefore, since there are not any audible music left; sound, once created, does not leave any residues behind. In this future society, where everybody seemingly listen to this “ultrasonic” music and enjoy it, nobody practically hear anything. This scenery inspired me to design a musical setting where the actors are seen as practically listening to nothing ,with their open ears, yet do react to music that is being played literally inside their heads through bone conduction!
Above, you can find some excerpts from J G Ballard’s science fiction story “Sound Sweep” with my illustrations. And here, you can view a more interactive narration of the story.

On the other hand, in an era where huge efforts are in the making for the democratization of musical creation and accessibility, it seems unfair to see hearing impaired individuals being a part of the communal aspect of music at a very restricted level. With today’s technology, we are able to cure most hearing deficincies, and the ones that are currently incurable (most kinds of sensorineural hearing loss) are being profoundly studied. Therefore, it is not the medical effort but the designer’s will what seems what is missing in enlarging hearing impaired’s social comfort zone today. The question I want to ask with this statement is: what can we do to create more inclusive spaces not only for those who are experiencing biological hardnesses (this is still a type of “pozitive” marginalization in fact) but to arrive to a level where these “impariment” implications are no longer relevant. In the search for an answer to this question, I decided to create a spatial music instrument to be played collectively by both hearing impaired individuals and individuals with optimal hearing.


To provide a bit background on the anatomic underlays of the hearing impariment, we can broadly state that there are mainly 2 kinds of hearing impairement: sensorineural hearing loss and conductive hearing loss. Today, if you are a hearing impaired individual and if your doctor is telling you that there is nothing the medical world can do, then you are most likely to have a sensorineural hearing loss which stems from deficiencies regarding the neural pathway of the audial perception. Contrarily, the conductive hearing loss occurs due to mechanical problems in the structure of the ear itself; before the soundwave captured by pinea reach cochlea and get transformed into a neural signal through the hair cells covering cochlea. In this type of hearing loss, as the problem is between cochlea and outer ear, bone conduction based hearing implants can be of solution since they directly excite the cochlea by by-passing the outer and the middle ear. If problem is in cochlea itself generally the deficiency can be resolved by cochlear implants.


In my project, I use a simple oscilator creating a magnetic field between two cones around which a string is tangled. As the string pass across cones, the change in magnetic field vibrates the string. This vibration is them amplified via a small metal piece connecting the string and to a large canopy covering cones. This metal canopy vibrates as the string vibrates. When one press this surface againts his skull, the bones constituting the skull starts vibrating, so do cochlea as an extention of these bones. This is how a piece of sound can reach to one’s cochlea without using one’s outer and middle ear.


Overall, since the project altogether was more ambitious than what is required by How to Make Almost Anything’s final project which is a demonstration of the combination of different skills learned during the course (hopefully in an interesting way), I decided to break the project into sub goals:

    • and build the dome
    • and build a bone conduction circuit that is embedded inside the dome
    • and build a LED circuit that flashes according to the beat of the audio
    • and build a vibration motor circuit that responds to the beat of the audio
    • 5.eventually, instead of using an audio file, design a gesture-based instrument so that the hearing impaired users can also produce the music itself other than just listening to it.

I aimed to complete the first 4 steps in the framework of this course, yet despite my efforts on trying to understand the way H-bridge works, I managed to do only the first 3. More on the process is below:


Part1_Fabricating the elements for the geodesic dome structure

A.Dome’s base

step1: Buying the metal tubes
I decided to use metal tubes for the base of the dome because this is where the vibration motors were going to be placed and steel is great to intensify vibration! I bought 10 of this EMT conduit from home depot; total cost was around $35.

step2: Cutting the metal tubes
I used the metal chop saw for cutting the tubes into the dimensions I needed in N-51


step3: Grinding the edges of the metal tubes
After cutting the metal tubes, the ends needed grinding.


For half of the tubes, I used metal grinder at N51, and for the rest, used the drilling machine with a grinder end.


step4: Bringing the tubes together to create the sitting elements of the dome

I had previously 3D printed the nodes:




Assembling the tubes with nodes:








et voila!


B. Dome’s struts

Struts are the main structural elements forming the dome. Their press-fit nature enables them to work as 3D conduits made of 2D sheets in a reversible manner which is essential for the modular and sectional nature of the project. besides,  having press-fit joints enables struts to be “openable” when needed, given the electric underlay of the project,  without disturbing their structural functions. My dream is for this structure to travel from one city to another by connecting to hearing impaired communities around the globe. In such a scenario, havin a 3D structure disassembling into 2D sheets offers an unbeatable logistic freedom!

step1: Buying the materials for struts
I decided to use 2 differet kind of material for the struts. One is a black museum board from Blick and the other one is a polycarbonate sheet from McMasterCarr. I bought approximately 10 black museum board (cost around $100) and 4 polycarbonate sheet (2 of 24×24″ + 2 of 24×48″; total cost of 4 boards were around $70).

step2: Making the cardboard press-fit struts
The press fit struts were to be made by cutting & scoring (with a laser cutter) then folding a black museum board. Therefore I started experimenting with different notch designs and scoring frequencies:


After deciding on the optimal design, since the frequency of my geodesic dome was 2, I only had to fabricate two different type of struts with different lenghts.




step3: Making the polycarbonate press-fit struts
the main reason why I wanted to make some of the struts with polycarbonate (which is clear yet not brittle as acrylic and not non-structural like polyprophelyn or acetat) is to stick some LED circuitry into its surface with vinyl cutter. However with polycarbonate, since it’s not possible to use the laser cutter due to the material’s release of toxic smoke, the process was more tricky than it was with cardboard. I was only able to use waterjet cutter for this material with which the concept of scoring doesn’t exist.

step4: Waterjet cutting the polycarbonate sheets






step5: Scoring the pieces manually by using a mask, cutter and hot air gun

So, I designed and lasercut a mask which helped me score the pieces neatly by hand.

after placing the mask on top of the waterjet cut polycarbonate piece, I clamped them to the desk to prevent layers from sliding.


then scored the polycarbonate carefully with a cutter. The mask helped me make sure that the lines were perfectly straight!


then, by applying heat to the fold lines, I increased the material’s elasticity which helped me fold it easily.



et voila!


Part2_Producing the electronics of the dome

A. Bone conduction

step 1: Deciding on the amplifier’s gain
I made 3 different iterations (gain:20, gain:50, gain:200) of my amplification board for bone conduction circuit. Used the datasheet for LM386:


gain 200
I connected a capacitor (20uf) between the pins 1 and 8 of the chip (LM386)




gain 50
I connected a capacitor (20uf) and a resistor(1.2 k ohm) between the pins 1 and 8 of the chip (LM386)




gain 20
pins 1 and 8 of the chip are disconnected




Finally, I decided on using the gain:20 one since the sound was the most clear in this circuit.

step 2: Fabricating 4 amplifier circuits


B. Beat reflective lights embedded on a clear strut

after vibrating the skull to create the sound within ones cochlea, I wanted to have a visual feedback as well as a tactile one. For the latter, I would need a vibration motor circuit and I decided to postpone it for future updates of this project since simply there was not enough time. However I decided to try the former.

step 1: Design & fabricate a small section of a LED strip for the proof of concept
homemade LED strip trace:


tracing copper sheets with vinyl cutter, trying different force intensities:


sticking them into a polycarbonate strut’s surface and populating the circuit with LEDs and resistors:


step 2: Connect the circuit to the beat detection system (arduino + processing)
Having made my own circuit for the core electronic part that was bone conduction, I decided to use an arduino duemilanove that I made during input devices week for the light addition. The programming environment was Arduino IDE and Processing (through running the standard firmata). I basically used the adapted version of my code from the interface programming week for beat detection on processing.

step 3: Actually fabricating 3 homemade LED strip embedded clear polycarbonate struts
After successing in the “proof of concept”, I decided to fabricate LED homemade LED strips for the lenght of at least 3 polycarbonate struts since I was separating the beats into 3 channels thru Processing. The main reason why I wanted to do my own LED strip by using a ton of LEDs and copper stickers is because I find the commercial LED strips super kitch. So, for the sake of being able to end up with exactly what I wanti I spent painstaking hours for the vinyl cutting of the copper, sticking the copper bands into the polycarbonate surface and soldering the resistors + LEDs into it. Ugh!!!


after vinyl cutting more copper bands, I started placing them onto the struts’ surfaces


then I soldered the resistors and LEDs


all looked pretty good, so I decided to try each strut with the actual beam detection current. Video below is a timelapse of my trouble shooting session

all done! now it was time to snap close the strut and articulate it to the structure!


C. Forehead band: interface between the bone conduction devices and the skull bones

step 1: Sewing velcro into elastic band
I bought 2 rolls of elastic band and a one roll of velcro. After cutting them into desired lenghts by making sure that they will be adjustable to different head diameters, I needed to sew the velcro into the elastic strap.


I had never used a sewing machine before but I learned how to use it pretty quickly. I was way faster compare to sewing by hand.Before using velcro and elastic band, I made some trials on a scrap textile.


Of, course I got it jammed couple of times but this problem can easily be solved by pulling the mingled thread out of the bobin repository (make sure to higher the needle before to not to break it!)


et voila!


step 2: Sewing bone conduction devices into the forehead band


finally all the velcro bands were sewed into the elastic band. now it was time to sew the bone conduction devices into the head band. I did that by hand since it required lots of attention.


now only thing I had to do was to solder the ground and VCC cables to the bone conduction device during the final assembly of the dome.


et voila!


Part3_Final integration and the set up of the dome


so far, it had been very challenging to come to this final phase. I appreciate to the truth in the illustration below:


The metal bases of the dome: check!

The black carton and clear polycarbonate struts: check!

Bone conduction amplifiers (4 of them!): check!

Adjustable forehead bands to interface the bone conduction devices to the skull: ckeck!

I have started with marking the decagon projection marking the edges of the horizonral struts. This reference is important to be able to know where to put the metal bases.


After placing the metal bases, I started articulating the black carton struts to them.



Then articulated the black carton struts to black carton struts.



even though the assebly theoretically looks fairly easy, it actually requires quite an effort since making holes in every branch of the connecting nodes is a painstaking process!


When all the struts were articulated, it was time to stretch the electric cables across them. The aim was to distribute sound (via an amplifier) from the sound source in the middle (for now, it’s a computer; at the end of the Spring semester, it will be a gesture controlled instrument!) to the each bone conduction devide embedded in a forehead band worn by an individual sitting on the metal base.


The reason why there are thousand cables is because I need 2 currents (ground abd VCC) for each LED strip embedded polycarbonate strut (there are 3 in total) and 2 currents (ground and VCC) for each bone conduction device (4 in total). Below, how the source is looking like:


and it was time to demo!

2 3 4 5 6 7 8 9 10

I told the users to plug their ears wittheir fingers to enhance the bone conduction effect!

Please stay tuned for the further iterations of the project, that is it for now!

Project 1 :: Grava (More Updates Due)


Grava is an adaptive chair that responds to the physical presence of the person sitting on it . It is based in generative design methods that utilizes material behavior based computation to allow for kinetic transformations as a passive effect.

The project would primarily investigate material behaviors and explore the possibility of using complaint mechanism within the framework of animating built environments.

Inspiration & Key Motivation

The prototype borrows from Frank Herbert’s ‘Chairdogs’ as described in “The Tactful Saboteur (1964)” and “Whipping Star (1970)”. The ChairDog as the name implies a hybrid between a pet dog and a lounge chair. The chair fits to the user and senses the users mood, and like a massage chair calms and makes the user comfortable. In Herbert’s universe the ChairDog is described in multiple stories as a ubiquitous object with multiple functional and responsive attributes :

The room’s standard model chairDogs had been well trained to comfort their masters, McKie noted. One of them nudged him behind the knees until he dropped his bag and took a reluctant seat. The chairdog began massaging his back. Obviously it had been instructed to make him comfortable while someone was summoned…
He indicated a chairdog against the wall to his right, snapped his fingers. The semi-sentient artifact glided to a position behind McKie. “Please be seated.”
McKie, his caution re alerted by Bolin’s reference to “uninhibited conversation,” sank into the chairdog, patting it until it assumed the contours he wanted.

One of the interesting attribute was that Herbert thought of the ChairDogs not as artifacts with multiple actuators but thought of them almost as if they had lifelike responsive qualities and moved without the use of motors or actuators.
The intention of embedding responsive attributes in our physical environments is a theme that has been used by a number of ther science fiction authors. J.G. Ballard concept of psychotropic house was based on a physical environment constructed using a new form of material plastex (a combination of plaster and latex) allowing for transformation and control of the interal shape of the house. The notion of physical objects physically adjusting and responding to humans can have numerous implications in the way living environments are designed and constructed.

…As I stepped forward, it jerked away, almost in alarm, the entrance retracting and sending a low shudder through the rest of the spheres……It’s always interesting to watch a psychotropic house try to adjust itself to strangers, ………

– From The Thousand Dreams of Stellavista, by J.G. Ballard.

Also Clifford Simak’s self adjusting furniture was an imposratant insoiration , with a direct impact on how Grava was conceived and designed.

The self-adjusting furniture, bought at a time when the management had considered throwing the hostelry open to the alien trade, had been out of date twenty years ago. But it still was there. It had been repainted, in soft and genteel pastels, its self-adjustment features still confined to human forms.

-From First He Died (Time and Again), by Clifford Simak.


The chair responds to the users weight and encloses the user with a soft shell when the user sits on it. It is intended to function as a personal ‘petting’ space , where the user can think and contemplate without external distraction. This privacy/ contemplation pod can be designed to respond and operate only to specific individuals based on weight calibration.

Other possible functions for this system can include
1. As a Gaming Chair with a mounted head display and suspended seating zone to augment the gaming experience .
2. Medical Applications – With the appendages having embedded body scanners
3. Super Massage Chair- integrate pneumatic appendages !
Technology Innovation

The chair is based on a process of topology optimization based complaint mechanism design . Complaint mechanisms are flexible mechanisms that can transfer input forces (such as the weight of a person sitting on the chair) at some other point to perform some usable output action (such as generating a sinusoidal motion on the chair surface ).

A system to design such mechanisms was developed in the Rhino3D/Grasshopper environment using a custom C# script . The process consists of the following steps :

1. Input two dimensional design space boundary
2. Define loads, supports and intended transformation sequences.
3. Generate topology optimization based complaint mechanism .
4. Extract contours based on stiffness gradients.
5. Define flexure hinges and rigid body material in the result.
6. Convert results for fabrication .

An interesting outcome of this process is the visual quality of the results. The skeletal outcome is very organic and reminds me of the HR Geiger’s designs or the chair that was designed for the Space Jockey

Project 2 :: Ivyoid (More Updates Due)

Untitled-1 copyPlantroid is augmented interactive plant pet developed for indoor environments. It is inspired by the the fictional scenario of artificial pets based in epidermits (as proposed by Stuart Karten) in Antonelli’s story “The Design Doyenne Defeats the Dullness”. Its basic setup consists of a living plant that is augmented by a robotic shell. The shell can sense the plants needs (moisture , sun , nutrient levels) and enable communication between the human owner and the plant. The Plantroid autonomously navigates around the house of its owner seeking areas with sunlight to perch in . They are low maintenance pets, seeking the owner’s attention when they need water or nutrients. They have artificial behaviors encoded in their interaction with the human owner such as swirling around with joy when they get the owners attention , or using social media to communicate their needs to the owner.

I developed a quick prototype ( the ‘deadliest’ weapon in a designers toolkit to paraphrase Antonelleii :)) and developed a light seeking robotic chassis on which a plant pod is mounted . Integrating a plant -need sensor into the equation now ( soil moisture , lux levels etc)

Physically-Suggestive Media

Self-Actuating Bridge


This project was inspired by a concept of automated matter which has taken form in many sci-fi films (X-Men, Watchmen, Matrix), and recently at the Media Lab. The idea is a hidden bridge that appears or materializes only when needed, enabling the user to confidently drive off a cliff or into the water, to be met by rising supports just in time. Ultimately the bridge supports shoudl support a car when present, and otherwise go away / retract / disentigrate.

X2: X-Men United - Magneto Escape

X-Men – Magneto Escape

Light Bridge from Halo

Light Bridge from Halo

Secret Garage Platform

Secret Garage Platform






InForm Actuated Tabletop

InForm – Actuated Tabletop

Preliminary Sketches

W1 01

Pneumatic Drawbridge 1.1


FE 01

Light – Photoresistor
Photoresistors are light-dependent resistor (LDR), also referred to as a “photocell”. Resistance across them decreases with exposure to light. Wired from an analog read pin to ground.Read from an analog read pin.

+ Easy to detect – The kind of light detected is the same wavelength we perceive naturally. Being able to easily see the input signal makes debugging significantly easier.
+ No additional hardware – In most demonstration conditions, light – in the form of daylight, synthetic fluorescents, LED, and halogens – is present

+ Inconsistent – Visible light levels vary A LOT between spaces. This makes for a bad output for a linear actuator.
+ Reacts to ambient light – In the context of a full-scale implementation, a photocell input would result in the bridge delaying for clouds, or any other opaque object that could move over the bridge path.
+ Reaction not limited to the car – Part of the appeal of a bridge that materializes out of nowhere is the fact that it only works under certain conditions, or for certain people.


FE 02Magnetic Field – Hall Effect
Hall effect vary the voltage drop from a power source according to the surrounding magnetic field- including the Earth’s resting magnetic force. THis allows it to sense up, down, and the presence of other electro-magnetic fields.

+ Reaction is limited to the presence of a magnet
+ Magnetic field detection would serve as a basic identifying factor that would enable only individuals cognizant of the requirements of the bridge to activate it, and do so discreetly

+ Magnets + electronics = ?x??1?$?
Other unknown side effects

Video Demo (clickthrough)How to Make - Final Project Development - Magnetic field detection


Pneumatic Actuation
Air powered movements. Using one air compressors, gates are turned on and off to inrease/decrease pressure in tubes and chambers to create movement.

+ Air is pretty friendly
+ Compatible with water
+ System could scale relatively well

+ Requires a large air compressor
+ Requires weights for submersion
+ Air compressor makes noise
+ Actuation makes loud pops with each pressure release
+ Lots of hosing requires
+ No small form factors for cool models
+ $$$

Hydrolic Actuation
Basically the same as pneumatic, but using liquids instead of air.

+ Pressure isn’t an issue.
+ Scales well
+ Very compatible with submersion

+ Requires bulky/large reinforced/structural material
+ $$$$
Magnetic Actuation
Magnetic field detection would serve as a basic identifying factor that would enable only individuals cognizant of the requirements of the bridge to activate it, and do so discreetly.

+ Reaction is limited to the presence of a magnet
+ Very little background noise
+ $

+ Magnets + electronics = ?x??1?$?
+ Side effects

Stepper Motor

Stepper Motor



Stepper Motor / Lead Screw Actuation
Oringally I didn’t find stepper motors appealing, but then I discovered these steppers with lead screws that act as linear actuators. These are used for many 3-axis machines and other various uses. They are also used in CD/DVD drives to move both the CD chassis, and the laser scanner. Because CD/DVD drives are now essentially obsolete, China and Russia end up buying lots of them as e-waste, recycle them for parts, and then sell them for SUPER cheap. So these stepper motors are plentiful, and cheap.

+ 1.7″ actuation distance + Lots of torque
+ Tiny
+ Cheap. Between $1-2

+ No datasheet
+ Tricky to wire to
+ Does not come with a slider


FSM0815-KD95 Specifications
+ Bipolar (2-phase) stepper motor
+ DC 5V
+ 1000RPM = 16.7 revs/s = 3.35cm/s actuation speed
+ 22 degree step angle
+ 16 pulse (pulses required to make one ful revolution)
+ Dimensions: 6×7.5×13.5mm / 0.24″ x 0.29″ x 0.53″ (L*W*H)
+ Screw Diameter : 2.5mm / 0.1″ Download My Model
+ Screw Length : 43mm / 1.7″
+ Weight: 33g

Stepper Sliders

FS 01

FS 03

FS 04

FS 05


Electrical Design

I decided to make the array out of discreet baords with no communication, mostly for the sake of keeping things simple. Eventually I would like to make a version with centralized controlled, as this would allow me to create more wholistic movments and animations with the pylons.

My board design was based off of the example found on this page. I added a jumper to the board for a remote sensor daughter board, and moved te power jumper around to accomodate a series of baord being powered in series by one cable.
FC 01FC 02

Autorouting in Eagle never gets old. Eventually I settled on this configuration.

FC 03FC 05

Circuits (Bulk Print)

Circuits (Bulk Print)

Exterior Cuts (Bulk Print)

Exterior Cuts (Bulk Print)

I also designed chips for a power jack, and the remote sensor for each primary board.

FC 12

Power Jack


Power Jack Circuits

Exterior Cut

Exterior Cut

FC 15

Sensor Board

Sensor Circuits

Sensor Circuits

Exterior Cut

Exterior Cut


FH 01 FH 02 FH 03


FF 01 FF 02 FF 03 FF 04 FF 05