DERIVING ART THROUGH LANDSCAPE
DERIVING ART THROUGH LANDSCAPE
Embedded Files
Robotic 3D painting
an exploration of morphing landscapes, where three-dimensional form informs terrain patterns creating detection feedback loops
with ABB-IRB120
Robotic art-making is an integrated process of art, computational design, robotic programming, and simulation. This robotic landscape exploration creates a live feedback loop between three-dimensional forms and their extraction of hidden geometries. The hand manipulation of the wire mesh and fabric casing directly informs the contour pattern, which is detected through a camera that senses and captures the point cloud data. Superimposing contours highlight the canvas’ complexity of steep valleys undulating between shallow terrain through three-dimensional painting.
The feedback loop between sensing the canvas form and using the Grasshopper contour extraction script live informs the project's geometric design. Each mock computational study and physical iteration has a unique geometric design specific to the morphology of the manually shaped form from which it is extracted. The depths and shallows of the morphed canvas result in gradients of contour linework that range from tightly concentrated lines representing steep slopes and spaced-out lines representing shallow landscapes.
hand molding process
hand molding process
The morphed wire mesh was overlaid with a stretched fabric to create a smooth surface for three-dimensional painting through the robot arm. For the final iteration, the lowest cavities were embroidered representing rivers running through valleys.
sensing and encoding
sensing and encoding
The contours are extracted with a grasshopper script containing the robot definition, actions, and code generation. The contour distance and direction are manipulated according to the desired visual outcome.
Mixed -media pipeline:
Mixed -media pipeline:
Sense:
Canvas placed under the IRB 120 with a peripheral sensor attachment that works as a depth-capturing camera. That captures the 3D geometry as a point cloud.
Encode:
The point cloud is internalized to create a smooth mesh for contour extraction. Through the 3d painting script, we get the code for simulation.
Simulate:
Created a module that allows for simulation and virtual testing in Robot Studio. After checking the script is safe to run.
Run:
Synchronized the module into the IRB120 for 3D painting
The final product is a mixed-media exploration of morphing landscapes, where the three-dimensional canvas form informs the terrain pattern. The output looks beyond pattern building as a form of art by exploring the creation of sensor-based feedback loops that cycle through human input, robot sensing, programming, simulation, and 3D painting with a robotic arm.
Date: October 2024 | Location: No siteType: Team project with Visha Ahmad, Tejal Kotkar | Professor: Josh Bard
Google Sites
Report abuse