Duncan A. Buell
NCR Professor of Computer Science and Engineering
Dept. of Computer Science and E
University of South Carolina
Columbia, SC 29208
2263 Innovation Center
Office hours: TBA
Elections and Voting Technology
A recent presentation on South Carolina's previous election system.
I have been placed on a panel of the U.S. Commission on Civil Rights regarding voting rights, to be held 22 July 2019 In Ft. Lauderdale, Florida.
The link to the PowerPoint for my presentation is here
the link to the narrative for my presentation is here.
A link to a paper submitted for publication that includes a longitudinal analysis of the South Carolina election system based on the published data.
A link to my analysis of the November 2018 election, done for the League of Women Voters of South Carolina.
Why is it that we ought not trust computers for voting? Well, one picture is worth perhaps 1011.43 words.
A link to computer security reports obtained by FOIA about the South Carolina election system in fall 2016 and spring 2017.
A link to national news about the South Carolina election system in fall 2016 and spring 2017.
A link to a Wall Street Journal article about the South Carolina election system in fall 2016 and spring 2017.
A link to a radio interview about the South Carolina election system in fall 2016 and spring 2017 and the WSJ article.
Ph.D.: Mathematics, University of Illinois--Chicago, 1976
M.A.: Mathematics, University of Michigan, Ann Arbor, 1972
B.S.: Mathematics, University of Arizona, Tucson, 1971
Electronic voting systems
Computational number theory
CSCE5355: Foundations of Computation (Spring2020)
CSCE548: Building Secure Software (Spring2020)
CSCE513: Computer Architecture (Fall2019)
CSCE522: Information Security Principles (Fall2019)
March For Science, 22 April 2017
Here is the text of my short speech at the Columbia SC March for Science.
I have to take this opportunity to put in a plug for my book,
Data Structures Using Java.
I have recently moved into the digital humanities. I am working with scholars on various projects, including a Web 2.0 project and an educational game for language and morpheme learning. We recently completed an education game for early modern British history.
|More significant, though, is that I am working with Dr. Heidi Rae Cooley of Media Arts/Film and Media Studies on a critical interactive, Ghosts of the Horseshoe, a mobile application to convey the scholarship of Dr. Robert Weyeneth in History on the relationship of the University of South Carolina (then called South Carolina College) and slavery in the years prior to the Civil War.|
I am also working with Dr. Cooley on the digital augmentation to her book.
Electronic Voting Systems
I have also been working since 2004 (details can be found here) with the League of Women Voters of South Carolina on the subject of electronic voting machines. I was one of a group of citizens who conducted an audit of the June 2010 primary and the November 2010 general elections, with some disturbing conclusions. This link discusses only what we have found in Richland County; we have found similar problems in nearly every other county we have audited. I am also a participant in the Election Verification Network.
Supporting Computer Science
Much of my time these days is spent on behalf of computer science, the discipline, by working in South Carolina and nationally to ensure that computer science is "marketed" properly to the public, especially in the K-12 school system, and that computer science "counts" toward high school graduation and university admission. Computer science is in fact "science" and should be recognized as such. I was from 2008-2012 the national university faculty representative to the board of the Computer Science Teachers Association and was part of the organization of a chapter of CSTA here in South Carolina.
|In the past my principal research interests were the algorithms and architectures for performing computations, such as those in discrete mathematics and text/string processing, for which traditional computer architectures oriented toward floating-point computations normally perform at a significantly reduced efficiency. One class of such problems are those computational problems in number theory for which fast integer arithmetic, often multiprecision arithmetic, is necessary. Another broad class of problems are those for which "custom computing machines" were devised. These include bit-oriented computations such as those done in string processing, and image processing computations, for which one often needs substantial parallelism, but for which 8-bit or 12-bit arithmetic is often sufficient accuracy.|
I was an instructor in mathematics at Carleton University in Canada (1976-77), a faculty member in computer science at Bowling Green State University (1977-1979) and Louisiana State University (1979-1985), and a member of the research staff at the Center for Computing Sciences of the Institute for Defense Analyses (1986-2000). Among other things at IDA/CCS I directed the Splash 2 project building a custom computing machine that used Xilinx FPGAs as the compute elements and whose applications were programmed in VHDL. Commercialized versions of and variations on Splash 2 are available from several vendors, including Annapolis Microsystems, Inc. (with whom I have no financial connections whatsoever). As part of popularizing the concept of custom computing machines, we created a conference, held annually since 1993, now called the IEEE Symposium on Field-Programmable Custom Computing Machines (www.fccm.org).
Some comments and links elsewhere
I am also concerned with U.S. education in the STEM fields (computer and information science; mathematics; natural sciences; and engineering and engineering technology) fields, and I have been involved as it seemed necessary for more than 25 years in defending science education against attacks from supporters of nonscience, pseudoscience, and the occasional downright fraudulent dishonesty. In that spirit I commend to the reader the award-winning website of the National Center for Science Education, the Talk dot Origins website and its award-winning subsite Panda's Thumb, the almost unbelievably clear prose of the Kitzmiller decision, the website information from the American Association for the Advancement of Science, the National Science Teachers Association, the National Academies, the Federation of American Societies for Experimental Biology, and from South Carolinians for Science Education. The science standards in South Carolina require that students learn how it is that scientists go about a "critical analysis" of that which is to be considered science. These websites can provide that information. As is indicated by the recent press release and position statement by the Interacademy Panel on International Issues representing many of the world's science academies, that critical analysis has clearly been done and the conclusions (at least at the level of K-12 science education) are clear as to what students need to learn.
I grew up and went to high school on the north shore of Lake Pontchartrain. Some of my family still live there, so I have a strong personal interest in the effects of Hurricane Katrina and the attempts to rebuild New Orleans in such a way that a colossal failure of "the system" could not happen again.
I have also found the Big Word Project to be amusing, so I bought the word "curmudgeon.''
In addition to the above, I update from time to time some other links of possible interest.
|Finally, let's get real. It is irresponsible as a technical person and morally reprehensible as a human being to advocate on behalf of nuclear energy unless and until we have shown ourselves capable as a nation and a society of dealing with the environmental fallout (no pun intended) of the use of nuclear power. Nuclear power is the environmentally most destructive of all methods of generating power because there is nothing to mitigate the hopeless physics of half-life and decay. We have, in the United States, demonstrated over the last several decades that we cannot sustain a policy with regard to nuclear waste for even as much as two presidential terms. Given that nuclear waste must be stored for somewhere between hundreds and tens of thousands of presidential terms, it becomes untenable to argue that we in any sense of the word "ought" to be using nuclear power as anything but a stopgap to be rid of as soon as possible.|