Skip to content

August 29, 2016

Robots and Entropy / Evan Daniel, Digital Media MFA 2017

by evandanielartist

One of the goals driving my work forward this summer was forming a dialogue between my own practice and the robotics lab.  Doing this in an authentic way is no easy task given the esoteric nature of my artistic practice: the core problem I’m addressing is memorization, specifically concerning the infinite, non-repeating number π.

In conversations with my cooperating supervisor at the lab, Dr. Bill Smart, we touched on the idea that memorizing π is a means of trying to be more like a machine.  There are significant tensions suggested by this idea.  Am I making myself more machine-like, or creating a humanized version of π?  Am I imitating the precision of a digital system, or is the process somehow machine-like?  These questions might be posed as a kind of more literal and modest version of Warhol’s “I want to be a machine.”  Bill and I spoke about means by which this might be represented through robotic movement and the idea that something is being conserved.  This led to a work called Entropy Conservation.  

In Entropy Conservation the randomness I generate by typing π is offset by the movements of a robot. I type π from memory into a laptop, which evaluates my progress and controls a PR2 robot to move with varying degrees of randomness. The more efficiently I recall π the simpler the movement of the arms, while slowness and inaccuracies lead to complexity. If the digits of π are considered random then, in a sense, entropy is conserved.


The PR2’s arms have 7 joints each, allowing for a a great deal of freedom when designing movement.


The PR2’s arms have 7 joints each, allowing for a a great deal of freedom when designing movement.


The PR2’s arms have 7 joints each, allowing for a a great deal of freedom when designing movement.

Screenshot from 2016-08-25 23-17-54

The PR2’s arms have 7 joints each, allowing for a a great deal of freedom when designing movement.

It was important to me that the robot was more than a means to an end, but an actor in the work roughly equal to me in presence.  That presence creates space for questioning the process through which this “conservation” is accomplished.  π and the robot are communicating through numerical representation (according to theorist Lev Manovitch this is a unifying “principle of new media”).  The robot’s movements, while dependent on the digits, are being controlled by the human data I’m generating by recalling π from memory, and which it then interprets.  But that data is limited.

Within the narrative of the work itself this implies a certain agency inherent in the robot.  Societally the implications of this might be significant.  For example, recent research by BF Malle suggests that people are more comfortable (in some contexts) seeing robots act in a utilitarian manner, which to my mind implies uncertainty about robot agency.  With this in mind, the nature of the exchange of data between humans and robots deserves serious attention within the field of human-robot interaction.

To assist with Constant Entropy I developed πGame, a simple, modular command line tool for recording information and displaying feedback when someone is recalling π from memory.  Since ROS uses independent Nodes, this can be used within different ROS projects to evaluate someone’s performance when recalling π and extracting data.


πGame loading a saved state

In keeping with my surroundings over the summer, I dedicated significant time to artistic research via experiments and short trials.  Many of these involved simulation tools within ROS.  At least one involved a somewhat novel control system: a Dance Dance Revolution pad.  I created a simple program (that can be used with ROS to control robots) that allows a user to recalls the digits of π through dance.  In its first iteration, though, the controller proved too unreliable to document.


I used this Dance Dance Revolution pad to input the digits of π from memory into a program

In addition to this I continued strands of my own practice that had less to do with robotics.  This included some painting as well as a series of drawings in which I “compiled” code by drawing it from memory.  To keep up with my other work please feel free to visit my site.

It’s hard to overstate the importance of programs like the Maharam Fellowship for broadening the scope of an artist’s understanding and practice in one short summer.  Concurrently, robotics labs (like this one) that value conceptual variety and input from traditionally separate fields are providing students with truly meaningful experiences.  For me, this was an experience at once new to my practice and perfectly willing to be in dialogue with it.  As such I can tell it will have a voice within my work for a long time to come.

%d bloggers like this: