The National Science Foundation, an independent federal agency devoted to researching and developing the sciences, recently granted $7.5 million to a group of five universities and one nonprofit research organization, enabling them to collectively study the reliability of electronic voting system.

The grant, announced on Aug. 15, will be used to fund a project called ACCURATE, or, A Center for Correct, Usable, Reliable, Auditable and Transparent Elections, said David Dill, professor of computer science and one of the participants in the project.

ACCURATE will be based at Johns Hopkins University, according to the Stanford News Service. Johns Hopkins University will receive $1.2 million over five years for the voting center, according to an Aug. 16 article in the Baltimore Sun. The remaining funds will go to the five other institutions.

The other participating universities are Stanford; UC Berkeley; the University of Iowa; and Rice University, and the nonprofit research organization involved in the effort is SRI International, based in Menlo Park.

Avi Rubin, professor of computer science at Johns Hopkins University and technical director at the University’s Information Security Institute, was chosen to head up the center.

The project focuses on making voting systems more reliable, Dill said. Dan Boneh, associate professor of computer science at Stanford, is also conducting research for ACCURATE.

“The research will target the fundamental barriers to making voting systems more trustworthy,� Dill told The Stanford Weekly. “Professor Boneh’s area of study is computer security and cryptography, while my area is formal verification — proving that systems do what they are supposed to do. These are both central to making systems trustworthy. Obviously, a system will not be trustworthy if someone has broken into it, or can read or fabricate data.�

Some of the experts conducting research for ACCURATE will study the psychological practicality of electronic voting systems by working with volunteers.

“The main application of psychology is usability,� Dill said. “A huge problem with voting technology has been voter errors and confusion. Many of the problems of Florida in the 2000 election, including the butterfly ballot, are usability problems. Those problems aren’t unique to Florida. They happen in every election with every type of equipment.�

The project has funds to last five years, which covers the 2008 presidential elections in the U.S. But participants in ACCURATE say their research may affect elections after 2008 more than the 2008 elections themselves.

“We might have some impact, but we’re conducting long-term research, and it takes quite a long time for new technology to make it into polling places,� Dill said.