_ i see the green light
tag: [work] [automavision] [phys]
date: 20230815
classification: Nero Chenxuan He Solo Exhibition installation
location: Galaxy Museum of Contemporary Art Chongqing China
curator: Jingbo Huang
This installation reimagines scaffolding as an adaptive architectural system produced through human-machine collaboration within a custom gaming interface. Rather than functioning as a fixed support structure, scaffolding is treated as a generative framework that responds to user input and computational logic in real time. Operating as a drawing machine, the game employs a nonlinear design process in which users co-design with the computer, allowing the scaffolding to fold, extend, and transform from a staircase configuration while continuously negotiating material distribution and structural efficiency.
In the digital environment, cameras establish visual connections that guide the system’s behavior and spatial organization. In the physical installation, these cameras are replaced by projectors that cast imagery through the scaffolding without producing shadows, merging visual data with material structure. The projector’s inherent instability introduces a condition of constant recalibration, requiring ongoing negotiation between human adjustment and machine output. Calibration is therefore not treated as a resolved interaction but as an evolving spatial dialogue in which design emerges through continuous feedback between perception, computation, and physical assembly.
The project further deploys Automavision as a constructional strategy rather than a representational effect. Projection light becomes a demanding operative tool: fabrication and assembly inevitably introduce human error that disrupts geometric precision, while the projected image exposes even minor discrepancies between computational geometry and physical construction. The apparent exactness of the digital model therefore encounters the tolerances, errors, and contingencies of material assembly.
This problem became particularly significant because the project was developed in the United States during the COVID-19 pandemic, when international travel was severely restricted, while the physical installation was constructed in China. The geographical separation between design and construction transformed communication itself into part of the experiment. A digital model could describe the installation with considerable precision, but precision in a drawing did not guarantee that the same computational logic could be understood, constructed, or evaluated remotely.
The difficulty became most apparent in translating a Boolean operation—a procedure native to digital modeling—into physical construction. Within the digital model, projected lightboxes Boolean the scaffolding, establishing spatial fields from which structural elements are effectively removed. Yet instructing a construction team to reproduce a “Boolean” provided little practical means of determining whether the assembled structure was correct. The computational operation therefore had to be translated into a physical rule that could be understood and verified on site: where the projected light reaches the wall, no shadow from the scaffolding should appear.
Shadow consequently becomes both construction error and quality-control device. If a structural member interrupts the projection and casts a shadow within the designated field, the scaffolding must be adjusted. If the projected light reaches the wall unobstructed, the operation has been successfully constructed. The physical installation does not need to be geometrically identical to its digital model; instead, it must reproduce the operative relationship encoded by that model.
This translation changes what a drawing is expected to do. Rather than describing solid geometry alone, the drawing communicates a relationship between the edge of projected light and the structural frame. Light functions as a Boolean operator—carving, aligning, and defining spatial limits through projection rather than material subtraction. An abstract computational command is thereby converted into something that can be seen, tested, constructed, and corrected physically.
The distance between designer and builder consequently becomes productive rather than merely restrictive. Remote construction exposes a fundamental problem within digital practice: computational precision does not automatically produce architectural precision. Information must survive translation between software, drawings, instructions, people, materials, and construction procedures. The installation therefore accepts discrepancy between its digital and physical states while demanding fidelity to the relationship that organizes them.
Within this process, error is not simply something to eliminate. Misalignment becomes visible through projection and initiates another cycle of observation and adjustment. The digital model instructs physical construction; physical construction alters the behavior of projected light; the resulting image reveals discrepancies; and those discrepancies produce further physical adjustments. The installation is therefore never the straightforward materialization of a predetermined digital object. It is continuously negotiated through feedback.
![]()
![]()
The difficulty became most apparent in translating a Boolean operation—a procedure native to digital modeling—into physical construction. Within the digital model, projected lightboxes Boolean the scaffolding, establishing spatial fields from which structural elements are effectively removed. Yet instructing a construction team to reproduce a “Boolean” provided little practical means of determining whether the assembled structure was correct. The computational operation therefore had to be translated into a physical rule that could be understood and verified on site: where the projected light reaches the wall, no shadow from the scaffolding should appear.
Shadow consequently becomes both construction error and quality-control device. If a structural member interrupts the projection and casts a shadow within the designated field, the scaffolding must be adjusted. If the projected light reaches the wall unobstructed, the operation has been successfully constructed. The physical installation does not need to be geometrically identical to its digital model; instead, it must reproduce the operative relationship encoded by that model.
This translation changes what a drawing is expected to do. Rather than describing solid geometry alone, the drawing communicates a relationship between the edge of projected light and the structural frame. Light functions as a Boolean operator—carving, aligning, and defining spatial limits through projection rather than material subtraction. An abstract computational command is thereby converted into something that can be seen, tested, constructed, and corrected physically.
The distance between designer and builder consequently becomes productive rather than merely restrictive. Remote construction exposes a fundamental problem within digital practice: computational precision does not automatically produce architectural precision. Information must survive translation between software, drawings, instructions, people, materials, and construction procedures. The installation therefore accepts discrepancy between its digital and physical states while demanding fidelity to the relationship that organizes them.
Within this process, error is not simply something to eliminate. Misalignment becomes visible through projection and initiates another cycle of observation and adjustment. The digital model instructs physical construction; physical construction alters the behavior of projected light; the resulting image reveals discrepancies; and those discrepancies produce further physical adjustments. The installation is therefore never the straightforward materialization of a predetermined digital object. It is continuously negotiated through feedback.
The game does not produce a singular optimized solution. Human decisions initiate transformations, while computational constraints redirect what can occur next. Design develops through this exchange between intention and procedure, allowing the scaffold to emerge incrementally rather than from a predetermined formal image. What is drawn is therefore not simply geometry, but a sequence of negotiations between human input, computational operation, and the constraints of physical construction.
Within the digital environment, cameras establish visual connections that guide the system’s spatial organization. In the physical installation, these cameras are replaced by projectors whose images pass through the scaffolding without producing shadows. The relationship between camera and projector establishes a reciprocal system between seeing and constructing: one organizes the digital environment through vision, while the other translates that organization back into physical space through light.
Within the digital environment, cameras establish visual connections that guide the system’s spatial organization. In the physical installation, these cameras are replaced by projectors whose images pass through the scaffolding without producing shadows. The relationship between camera and projector establishes a reciprocal system between seeing and constructing: one organizes the digital environment through vision, while the other translates that organization back into physical space through light.
The distance between designer and builder consequently becomes productive rather than merely restrictive. Remote construction exposes a fundamental problem within digital practice: computational precision does not automatically produce architectural precision. Information must survive translation between software, drawings, instructions, people, materials, and construction procedures. The installation therefore accepts discrepancy between its digital and physical states while demanding fidelity to the relationship that organizes them.
Within this process, error is not simply something to eliminate. Misalignment becomes visible through projection and initiates another cycle of observation and adjustment. The digital model instructs physical construction; physical construction alters the behavior of projected light; the resulting image reveals discrepancies; and those discrepancies produce further physical adjustments. The installation is therefore never the straightforward materialization of a predetermined digital object. It is continuously negotiated through feedback.
I See the Green Light ultimately positions Automavision as a method for designing through perception, translation, error, and calibration rather than fixed form. Neither the computer nor the designer maintains complete control, and the constructed object is not required to reproduce the digital model literally. Architecture instead emerges through their incomplete correspondence—through repeated acts of seeing, instructing, constructing, checking, misaligning, and adjusting.
Within this process, error is not simply something to eliminate. Misalignment becomes visible through projection and initiates another cycle of observation and adjustment. The digital model instructs physical construction; physical construction alters the behavior of projected light; the resulting image reveals discrepancies; and those discrepancies produce further physical adjustments. The installation is therefore never the straightforward materialization of a predetermined digital object. It is continuously negotiated through feedback.
I See the Green Light ultimately positions Automavision as a method for designing through perception, translation, error, and calibration rather than fixed form. Neither the computer nor the designer maintains complete control, and the constructed object is not required to reproduce the digital model literally. Architecture instead emerges through their incomplete correspondence—through repeated acts of seeing, instructing, constructing, checking, misaligning, and adjusting.
