W Randolph Franklin

Electrical, Computer, and Systems Engineering

W. Randolph Franklin  develops and implements fast parallel algorithms on very large geometric datasets in CAD and GIS.   He's also modeled and processed large terrain databases, e.g., to compress, to compute hydrography and visibility, and to site observers, and compressed 5D environmental data sets.    The algorithms are the fastest in their class.

His current research hobby is designing and implementing small, simple, and fast data structures and algorithms for large geometric datasets.  Note that efficiency in both space and time can become more important as machines get faster.  This research is applicable to computational cartography, computer graphics, computational geometry, and geographic information science.

He is now currently studying parallel CAD operations, such as overlaying 3D triangulations (tetrahedrulations).  3D-EPUG-Overlay is an algorithm for exactly computing the intersection (it can also compute general overlays) of pairs of 3D triangulated meshes. Each input mesh may represent a single solid or a more complex polygonal subdivision (representing multiple solids, each one identified by a label). 3D-EPUG-Overlay employs a combination of several techniques to ensure its correctness and efficiency.

His  last major NSF research project was to understand the mathematics of terrain.   Another recent NSF project, together with Cutler and Zimmie, was analyzing how levees erode when overtopped, and testing models in a geotechnical centrifuge.  Subprojects include compressing terrain while preserving slope, compressing multidimensional environmental data, and extending partial hydrological data to a complete and consistent river network.

Another major project compressed and operated on terrain for DARPA/DSO/GeoStar.  The operations included multiple observer siting and path planning.  His longterm uneachable goal is to establish terrain modeling
on a proper formal foundation that respects its physical properties.

Previous positions include: Visitor at Georgia Tech, spring 2016, Program Director, Numeric, Symbolic, and Geometric Computation Program, CISE, National Science Foundation, 2000—2002, visiting positions at Genoa, Laval, CSIRO Canberra, National University of Singapore, 1992—1993, and Visiting Professor, UC Berkeley,

Focus Area

efficient parallel and GPU algorithms for large geometric and spatial datasets