Archive for the ‘ITP Thesis’ Category

More Art Robotica

Wednesday, May 20th, 2009

These are some of the Paintings that I made with Jackoon at the ITP Spring 2009 show:

Reference image, wasp on a flower:


Collaborative Painting:

Reference image, rodent a skull of a species in the order Carnivora:

Jackoon Painting:

As you can see on the painting of the Skull, Jackoon managed o paint the general area where the skull should be. The brush strokes where not as precise as should be, because the lighting conditions where not optimal at the ITP show so the tracking system was a little off.

Someone at the ITP show told me something really inspiring; “This is the worst work that [Jackoon] and the other artbots will paint. They will only get better.” I agree, even now the paintings are amazing.

Link to pieces that where hanging at the show while Jackoon was making the art above.

Art Robotica in the ITP Spring Show 2009

Saturday, May 9th, 2009

This is some of the art that I showed at the ITP Spring Show 2009 exhibit.

The Art:

“Lemon Wasp” 2009
Acrylic on paper, 48”x 48”

“Daisy” 2009
Acrylic on paper, 48”x 40”

The Art Bots:

Jackoon at work:

Jackoon Painting from Oscar G. Torres on Vimeo.

Art Robotica: Thesis Presentation!

Tuesday, May 5th, 2009

Art Robotica is a term used to describe the body of work created by the collaboration of “ArtBot” and artist. ArtBots are Autonomous Machines that paint or draw on a canvas, sheet of paper or a surface of some kind. Art Robotica reflects our new codependency and symbiotic relationship with machines. ArtBots are, in essence, collaborators rather than just new tools.

A presentation of Art Robotica was part of ITP Thesis Week 2009.
The thesis presentation was Tuesday May 5, 2009 at NYU’s ITP (721 Broadway, 4th floor)

Web Video Archive:

This is one of the paintings created so far:

“Girl” 2009, Acrylic on paper, 48″ x40″

To find out more about Art Robotica visit:

More Art Robotica:

Artbots The Testing Begins

Monday, April 6th, 2009

The ArtBots begins to take shape. I started the project by tackling the bot that will be the most complex, the hexapod.

I made a the basic form of the bot using some wood screws and tape to begin testing the movement, software and circuit.

Servo Motor Test from Oscar G. Torres on Vimeo.

I was originally going to attempt to use an Arduino Micro controller to do all the controlling of the servos (video above.) But I slowly realized that to control the timing of all the servos at different time intervals will be a big pain in the butt since the “delay()” function messes with the Arduino code enough to get really buggy.

To hopefully save a headache I bought some servo controllers made by Pololu And so far with the help from Xiaoyang (Alex) Feng and the info posted online about the controller, I got the results I was looking for.

The Pololu Code I used to make the servos move can be found on the Pololu forums The one things that they did not include which is a little confusing, is how to set up the servo controllers if you plan to daisy chain them to control more than 8 servo motors.

To start you should set up your Arduino (above) and a breadboard with the Pololu servo controller (below.)

A couple of things to note:
1. Two different power sources are needed for the servo controllers. One for the controller’s micro controller and the other to power the servo motors. Both power sources should have a common ground.
2. When you set the controller’s number setting, you need to do one at a time.

This is the code I used to program my second controller to know its servos are numbered 8-15 and not 0-7 as they are programmed by default.

The parameter that sets the servo numbers is the fourth “soft.Serial” line of code “softSerial.print(0×01,BYTE);” This variable can be either 0×00 (hexadecimal) to set the controller to use number 0-7 for the servos, 0×01 to set the controller to use number 8-15 for the servos, 0×02 for numbers 16-23, etc…

Once you upload this code you must wait for the servo controller to get reprogrammed. You will know the servo controller’s settings where changed by the blinky lights on the controller.

Pololu Servo Controller Settings from Oscar G. Torres on Vimeo.

Once the code has been uploaded and the settings where changes, you MUST delete or comment out, the three lines that set up the servo numbers since you only need to do that step once.

the lines you need to delete or comment out are the following:

Once you comment out the code above, re-load the code onto the Arduino.
When the code is finished uploading nothing will happen, because you need to re-start the Pololu servo controller for the settings to kick in. You can simply unplug the Arduino and the plug it back in to refresh everything. The servos should now move a few degrees back and fourth. Don’t forget to power the servos separately and to have a common ground.

Servo Controller Test from Oscar G. Torres on Vimeo.

I daisy chained three servo controllers together so that I can control 20 servo motors. This is my first test, one serve per servo controller. So far so good.

Testing Servo Controller with Three Servos from Oscar G. Torres on Vimeo.

Oscar’s Thesis Concept: The ArtBot

Monday, March 9th, 2009


ArtBots are painting machines that paint their environment in a unique artistic style. They collaborate with each other and with an artist to make a painting. Each robot will have a personality which will translate to their brush strokes and painting style.

These are two quotes from a couple of my favorite artists that inspired me to pursue the concept of a robot artist:

“[painting] is colored dirt smeared on a flat surface, usually stretched around some wooden sticks. And it is for me, the most magical of all mediums. You smear that colored dirt and it makes space where there is no space” ~Chuck Close

“The photograph meant the death of painting, but its resulting to be the opposite. Thanks to photography, art is resuscitating.” ~Salvador Dali

This is what I think the robots will look like in a gallery setting:


The way the robots will function is similar to the way that Jackoon functioned. There will be a video camera in the ceiling that will track the movement of the robots and will be used to know when the robots should paint. All the logic for the painting composition and collision detection will be done in a laptop and sent to the robots with wireless communication. This is an image of the set up:

The mechanics for the mobility of a couple of the robots are inspired by Theo Jansen’s leg design. The other robots will have servo motors and are inspired by hexapod.

I am not the only one using Theo’s leg design, I have a feeling that this design will become very popular as time goes by.

This is a rough sketch of what the robot using the Theo’s leg design may look like, the arrows point to the direction they are facing:

And this is a robot who’s leg design is similar to Theo’s concept but I am hoping that I will be able to get better movement from this design.

This is a simplified sketch of the isolated legs and the places where joints exist:

These are the physical legs that I built and am in the process of testing and some videos that I took of the rough motion that is achieved by the respective designs:


Theo Jansen Leg Design from Oscar G. Torres on Vimeo.


Theo Jansen Inspired Leg from Oscar G. Torres on Vimeo.

Stay tuned for updates on the ArtBots!

Initial Thesis Concepts

Wednesday, February 18th, 2009

Solar Rail system for a train was a concept that I wanted to develop for thesis, but after thinking about the scope and time line of the project, it became cleat that I was being to ambitious. Getting a working prototype will require a little more time and research. The basic idea was to make a network of solar modules that collect energy onto giant capacitors or batteries or some other form of energy storage. The energy would remain stored until the train physically touched the module and began to drain the energy. The result will be a train that gets propelled by an impulse system. This is a rendering of a section of the rail:


An other idea I had was a little rover similar to the Mars Rovers that NASA used to explore Mars. The rovers I wanted to create would communicate with a wireless WIFI connection and be controlled remotely via a browser interface or mobile interface. The drawback from this idea is that there is a similar bot in existence called the rovio so this deterred me from going on this direction. These are the initial sketches:


The next idea I have in mind is to expand on the concept of a painting machine. Last year I made Jackoon the painting Robot. Since that project worked so well, I will be adding to it and making painting bots that are more flexible and can work together to make a painting. This is a quick sketch of the bot:


This is a rendering of what I think the bot will look like:


I will be writing more about the ArtBots soon, once I can wrap my head around what I really want out of these artists.