Max Chandler  Robot Art

FAQ



Art With Robots
Asian Art Influences
Meet The Robots



Art With Robots


Why use robots to make art?

Our time is different than before because we have smart devices, such as TV remotes, cell phones, game machines, computers, ... all around us. They are part of our lives. Art that is really about us now should acknowledge what it means to be human now. For me, using robots allows me to combine my art and technology backgrounds into a whole.


What is the robot's role?

In my view, the robot should be a tool that concentrates on mark making. This gives a new and vital look that can not be achieved in another way. I have several robots each designed to make marks in a variety of ways. See Meet The Robots for more information about each robot.


What is the artist's role?

The artist creates both the robots and the artworks. Depending on which robot and which methodology is used, a variety of media can be used. Currently, using robots, I have made pen and ink drawings, pastels, oil paintings, and acrylic paintings. See Recent Work for my most current work. The artist should be making all of the decisions about the art, from designing the robots, choosing the palette, composition, content, structure, etc. The way I think about robots is as a new and exciting kind of brush.


What is your methodology?

I have a number of different strategies, but they all fit into a pattern that begins with observation, then analysis, and finally execution. Each of these stages contains a fundamental conflict that needs to be resolved in a unique way for each artwork. In observation, which I mostly do with traditional drawing, there is a conflict between actual vision and knowledge. A lot has been written about these issues. Mostly the literature on drawing wants the artists to become a pure observer and get into a mode (sometimes called right brain) that concentrates on hand/eye coordination rather than knowledge. This is a major over-simplification. This substitutes knowledge of edges, for knowledge of the object and turns the artist into some kind of camera. Human vision is binocular and from multiple viewpoints. The real challenge is to make an informed results that shows knowledge of structure, forms, light, growth patterns, and much more.

In analysis, the fundamental conflict is between the parts and the whole. The stage I call analysis includes composition but adds in how to get the robot to work and participate. Most of my current robots have vision sensors or multiple vision sensors to determine how they will turn and travel. I have several strategies for designing images that work with this type of robot. I am also developing more types of robots that will expand what I can do. See the FAQ question How do you instruct the robots how to paint? for more detail. This stage of the art involves the most thinking, but it is important to use intuition and feeling. This stage is about balance and the results is a plan, usually with masks and templates and notes.

The execution stage is very much like a performance. I mix all the paints, select brushes, etc. ahead of an execution session. I place the structural templates and start the robot. I design the structure so that the robot moves on and off the surface. I use the off surface times to change brushes, reload colors, set different controls to make the actual marks. I mostly use multiple brushes so that the paints will blend on the canvas. Depending on which robot, I must choose which brushes, colors, attachment methods, etc. to begin the mark making. Some of the robots are slow and careful, some are very fast and nervous. These sessions are exciting times with a lot of intense action. To make a single work will take ten or more sessions. Each session becomes a layer that builds the interaction of the marks and colors into a beautiful image. The fundamental conflict in the execution stage is between plan and chance. If the robot and the plan were perfect, the actual image would match the plan. Many things happen when an real robot is moving an real paint brush on an real canvas that will not be like the plan. The wheels slip, some turns are smaller than the turning radius, the template is incomplete, etc. This plan versus actual is how the art becomes as much a product of the robot as well as the artist. In the final layers of execution, I remove the templates and let the robots marks become the guide.

Someone once asked me if I was ever surprised by the final work. My answer is -- always. I deliberately make structures and plans that are at the edge of what the robots can do. I believe the most beautiful results occur when chance has a role. This also means that many of the paintings I make are destined to be painted out, and simply be part of learning how to make a better image next time.


Has anyone made art with robots before?

Yes, quite a few

The first solid example I know of is Seymour Papert. His initial work with Logo for teaching children had a "turtle" robot that drew from language commands. Here is a short article that shows this work. The electromechanical turtle was quickly replaced with a cybernetic (computer cursor, in this case) turtle that became the center of Logo's turtle graphics. The art work wasn't spectacular, but the benefits for children and education certainly were. You should search for Seymour Papert, if you don't know about this wonderful man.

Howard Cohen has been the most serious artist to work with robots. Here is a review of a book about Cohen and his robot Aaron. His approach depends heavily on Artificial Intelligence. This was an interesting and large project that Mr. Cohen pursued for many years. The technology has changed a lot since Mr. Cohen's Aaron. Today we can make robots that are very much like the smart devices, you use every day. My art is involves robots like the smart devices.

Alan Rath has made some very good kinetic robot sculptures.


What are some types of art with robots?

Drawings and paintings -- which is what I do

Kinetic Sculptures like Alan Rath and several others

Static Sculptures that are imaginary robots, not actual working devices.

Most robots are from literature and entertainment. Robots existed in plays and novels long, long before there were actual robots. Actual robots are almost nothing like fictional robots.

Many remote control devices are incorrectly called robots and some robotic art is just electronic puppets.

top


Asian Art Influences




Why do your paintings look like Asian art?
I try very hard to have my art look like Asian art. I have always liked Asian art, since I was a child. Two of my most favorite art references were The Mustard Seed Garden Manual and Japanese Aesthetics when I was young. Later in life, I went to graduate school in Taiwan and studied Chinese painting and calligraphy. The contemporary art in Taiwan was very strong in the 60's with good artists like Chen Ting-shih and Liu Guo-sung showing how to stay traditional and be modern at the same time.

The painters I thought were the strongest included Chinese Zen painters like BaDa Shanren and the Ming Dynasty Wu School. The calligraphers I studied most are Yen Chen-ching and Yu You-ren. I had the opportunity to see many Zhang Da-chyan paintings, as well.

top


Meet The Robots




What kind of robots do you make?

I use robots from several sources.

Legos Mindstorms, especially with tehnics and what are called the rugged construction techniques. Many legos robots are fragile and easily break. That won't do for making serious art. By using the rugged constructed as described by Fred Martin at MIT, my robots have run for hundreds of hours without breaking. Legos are used to teach robots at MIT, Carnegie-Mellon and many other universities. Legos are use for robot prototyping many places including NASA. Because I only need one of a certain type of robot Legos are a good source. Also Legos Mindstorms community is very large andthere are excellent programming resources available. I use the leJos system and the Java language.

Robots have usually been divided into two main parts. Brains, for the microprocessors and sensors -- Brawn for the mechanical and electromechanical parts. Besides legos, I use the Java version of the Basic Stamp (called javelin) for brains and the BOE series for brawn.

My robots fall into two main categories: walking or wheels.

Walking robots need to shift the center of gravity for each step, the COGS(center of gravity shift) makes for interesting marks. I call my walking robots, Gimpy, so far I have Gimpy1, Gimpy2, and I just assembled Gimpy3 --more to come. These robots use light and/or color sensors to follow templates that I construct for them.

Because the wheeled robots are more accurate, I use graphics languages to direct them. The wheel series of robots are called Zeb and their language I call Zeb GL. It is like Open GL, but very much smaller. As I build robots with bigger and bigger brains Zeb GL will grow. One of the wheeled robots can make a drawing from a midi file by translating the notes into movements.


How do you instruct the robots how to paint? I am constantly working on ways to improve and refine these methods. Here is a sample of some:

Line Seeding. With this approach, I draw a small line fragment, like a dash or an "x", arc, etc. I start the robot looking at that segment with heavy enough markers(pastels, also) so that the robot will quickly see its own marks. The image evolves from the robot's reaction to the line seed.

Shape Following. This is similar to line following, which is a common robot path control technique. Line following usually tries to see both sides of the line and straddle the line. This would not be very good for making pictures. Shape following aligns with one edge so it can travel around a shape. I make shaped templates that are used and reused to guide the robots.

Store Patterns. I can also store the x,y points of the outline of a shape in the robots memory and have the robot trace and/or fill these patterns. This is the most common way I build the cactus in my paintings. top