Computational Design Symposium @ CMU

If these files could talk: Extracting Patterns from Digital Archives of CAD Drawings and 3D Models Using Novel Forensic Techniques


What can we learn from examining a given digital file of a drawing or 3D model? We could certainly open up the file in the appropriate software program and examine the drawing or model itself, but what more is hiding beneath this virtual representation? While a drawing or 3D model file appears to us on screen as collection of parts existing at one and the same time in virtual space, the constituent lines, triangles or surfaces are actually represented in the software program as a sequence known as an index. By examining a file’s element index, information can be gleaned about the designer’s drawing or modeling process (the process of a file’s becoming) and about the idiosyncrasies of the software program in which the file was created.

Just as archaeologists have found hidden paintings concealed beneath famous works of art using radiography, is it possible to uncover another layer of information in our digital files lying just beneath the surface? Finding strategies of representing the sequence of elements in an index to uncover patterns is the subject of my research. This work requires building a toolbox of forensic methods to distinguish between user or algorithm-derived indexes across various file types and software platforms.