In the performing arts--specifically dance, opera, and theatre--creative personnel routinely collaborate with their technical colleagues in the development of the physical environment constructed and experienced on the stage. In this instance, choreographer Yacov Sharir and scenic designer Richard Isackes envisioned a dynamic, translucent 3-dimensional projection surface as a primary scenic element for an original dance-theatre piece titled Automated Body to be performed at the University of Texas at Austin. The technical challenge, to motorize and automated the component pieces of this stage sculpture, was undertaken by the Professor Fritz Schwentker with the assistance of graduate students in UT’s Department of Theatre and Dance.

The scenic effect in question consists of a series of fine clear-acrylic beaded curtains suspended from 3-foot x 3-foot frames [see Figure 1 below].

Figure 1: The "Box" with dancers in performance.

Nine of these curtain frames arranged in a 3 x 3 array comprise a box nine feet on a side and ten feet tall. The clear-acrylic bead chain serves not only as a curtain, but as a projection surface for the lighting effects, video, and computer generated graphics seen throughout the production. Since the bead chain that makes up the curtains hangs on one-inch centers, it is permeable in a manner that allows performers to move inside the projection surface. But, the creative team for the Automated Body Project needed the suspended pieces to move vertically within the performance space at various times throughout the piece--sometimes simply to fly the box clear of the stage floor itself, but more often to arrange the component parts of the box’s array dynamically as part of the choreography. The ability to control the motion of each section in terms of both speed and position proved crucial to offering the variety of arrangements the choreographer and designer might want on stage as in Figure 2.

Figure 2: The "box" moves to assume various shapes.

With the above in mind, the mechanical design called for nine cable-drum winches to move individual curtain-frames [Figure 3]. A modest budget for equipment dictated that each of these mount an AC three-phase induction motor fed by a variable frequency vector drive. The control for these nine drive/winch units needed to allow for the broadest possible range of speed and position combinations. Even more crucially, the control had to allow those combinations to be stored easily during the technical rehearsal process so that they could be replayed accurately during performances. The operator interface had to allow accurate visual feedback of the motion and position as many moves happen in darkness or in views obscure to the stagehand/operator. Finally, the interface had to be safe and simple to use—because, as is typical in live performance, there is little time for operator training, and the technical rehearsal period in which all lighting, audio, automation, and choreography are brought together with the performers is compressed, hectic, and often exhausting for all participants.

Figure 3: Typical cable winch.

The control system for the Automated Body uses two Flexmotion cards (one four-axis and one six-axis version) and their associated wiring interfaces to serve the nine vector drive axes. In order to offer excellent speed control and synchronization, the Flexmotion controllers and vector drives are treated as servo systems in this application with position feedback supplied by shaft-mounted incremental encoders on the winch frames. Flexmotion VIs provide convenient access to motion parameters needed by the controllers; these are called by an HMI built and running under LabVIEW on a standard desktop PC.

The operator interface for the Automated Body project as seen in Figure 4 performs three major functions—it

1) provides an input method for trajectory data

2) offers visual feedback on axis positions, and

3) gives access to the cue-sheet.

In fact it is the requirements of this last component that drives the structure of the entire process. Using the paradigm of a cue and cue sheets familiar to theatre practitioners for generations, any motion cue is comprised of the following essential data for each axis: whether the axis is selected to run, target position, top speed, acceleration, and deceleration.

To LabVIEW, the motion data for each axis is a cluster of the five essential components noted above, and a cue is an array of clusters, its size determined by the number of axes available or needed for the performance in question. A cue-sheet then is an array of cue arrays, the size of which is also user determinable. The operator can call up a blank cue-sheet or one previously saved to disk; as long as a cue-sheet is loaded, the cues can be read or saved during technical rehearsals or performances at any time as required through calls to sub-VIs. While the interface is running, these sub-VIs access global variables holding the currently active cue, the cue-sheet, and flags noting whether a cue has been changed or saved. In this project, LabVIEW’s cluster and array tools made possible the development of simple methods to update motion data in the axis clusters globally (e.g. speed, acceleration, deceleration.) rather than the operator setting each parameter individually.

Figure 4: Front panel interface for the Automated Body controller. In addition to setting position and speed for each axis the operator can enable the axes for each cue.