Can Robots Dance?

Industrial robots have become instrumental to our way of life. Fascinating…but can they dance?

Industrial Improvisation, the thesis project by Chryssa Varna of the Interactive Architecture Studio–Research Cluster 3 at the Bartlett School of Architecture UCL (University College of London), is asking just this. In collaboration with UCL Lecturer in Interactive Architecture, Ruairi Glynn, and fellow PhD candidate and artist Ollie Palmer, Varna is employing contemporary dance as a surprisingly poignant platform on which to demonstrate the complexity of movement achievable when kinetic design and industrial robotics merge.

The creator, who unsurprisingly possesses a background in dance and photography in addition to architecture and design, is especially interested in merging the seemingly parallel universes of choreography and kinetic architecture--using technology to transform performance and stage design into living art.

In the study, a human dancer performs both pre-choreographed and improvised dances on a stage, engaging in a gestural dialogue with two elegant robotic entities suspended from the ceiling.

Industrial Improvisation from Chryssa Varna.

Using controversial 20th century choreographer Rudolf von Laban’s four part description of movement, the project inspects: BODY: As the different parts of the dancer’s anatomy move, the robot reciprocates, using the computer vision and body movement tracking powered by Kinect technology. TIME: Movements of both the human and robotic components occur in real time at different rates. DYNAMICS: The robotics react to different gestures with different qualities of movement. For example, a geometrical form created by the human dancer will yield an L-shaped movement from the machines, while more dimensional movements will yield a twisting response. Thus, the interchange of movement remains dynamic and spontaneous throughout. SPACE:  The robotic fans and armatures constantly transform the space with different modulations, increasing and decreasing the spatial qualities between the dancer, themselves, and space itself. The next level of space alteration involves the installation’s observers, who take part by occupying the space.

To design the ultra-sensitive machines, Varna and company used three layers of human improvisation/interaction with choreographer Wayne McGregor’s processes in mind:

The first layer isolated 10 particular moves of choreography and added other spatial concerns to these movements to aid visualization of 2D objects and their properties (i.e. color, shape, texture, and sound). The second layer created mechanical and drawing gestures that would break down the fragmented forms and movements made by the human body. The third layer inspected and added a 3D aspect to the machines’ motions by taking measurements of the human’s movements and the perceived distance from the center, or torso.

The UR10 robots, made of six joints and two aluminum tubes, were then connected to cylindrical bases around which the machine would have a “workspace” of 1300 millimeters all around. Designers had to be aware to calculate the right distance of the robot from the base that would produce reaction rates that would most accurately mirror the speed of the human dancer’s movements. The finished product is as elegant and delicate as one could expect a robot to be.

The most exciting part of all this work is seeing how dancer and robots seamlessly alternate between the roles of choreographer and performer; creating and interacting even as they move and obey. In other words, the “gestural vocabulary” creates a substantial dialogue between human and machine; arguably, a true exchange of ideas. Ultimately, this beautiful experiment leaves observers with the question of whether kinetic design rendered in industrial robotics could ever perfectly match up to human interactions. Or on second thought, should we be more open to asking a cyborg to fance?