Robo_Craft. Robotic drawing
Ionuț Anton, Dana Tănase
For this year’s Renault Romania Day, we were invited to take part of the Dacia Research Divisions exhibition, showcasing the possibilities of creative use of industrial robots. For this exhibition we developed a custom application for robotic drawings, as part of our Robo_Craft research project.
For the Renault Romania Day 2014 we had the opportunity to develop a application that would use industrial robots in a new and fun way, as opposed to performing the same dull and mechanistic task each time. We were very much inspired by the work of Sigrid Brell-Cokcan and Johannes Braumann (Robots in Architecture),Sci-Arc Robot House and Richard Dank(TuGratz) that developed similar applications, but we wanted to give it a special twist.
We started evaluating several strategies for drawing generation and soon we focused on a Pointillism style inspired drawings that could be very well implemented. Some of the researched drawing strategies involved a grid of points sampling over an image, a random reduction distribution of points, a random sampling of points based on image brightness zones, a weighted voronoi sampling of points or even a variable line width drawing. From the point based strategies we then developed several finishes based on small strokes (pointillist style), hatches or delaunay diagrams.
Using a series of base images we then perfected the workflow and selected two strategies that we implemented in a final Grasshopper Script that would run for the event day. During the Renault Romania Day we had pictures taken from the visiting public that were used as a base for the algorithm and the robot fabricated the portraits. As the event was focused on families and children we had an enormous success.
Some credits for our inspiration and open source software that we incorporated in our own workflow is required:
To Evil Mad Scientist for the StippleGen code done in processing that was a starting point and inspired the random reduction sampling of points.
To Sahab Yazdani for the Weighted Voronoi Stippling executable that we used for some point generation.
To Peter Vikar for the good documentation of the MisPortraits series.
To Francesco di Niccolo for a very nice algorithm for zoning images into tones and random point distribution.