Election problems generate opportunities

The U.S. presidential election is over. President George W. Bush took office following his inauguration on Jan. 20, but many of the questions surrounding the election are not entirely resolved. I am not talking about the political games that may or may not have taken place, but about questions relating to the mechanics of the voting process. As engineers, we cannot solve the political problems. We can only give technical input and provide technical solutions where they are available.

In Florida, the difference in the vote for the top two presidential candidates, as counted and certified by the Florida State Elections Board, was 156 votes, or just under 26 parts in one million (26:1,000,000). At the same time, there were several thousand votes that were never counted for a variety of reasons.

The vast majority of the uncounted votes were the result of the automatic voting equipment being unable to determine what the voter did. Thus, Americans have a situation where we do not and can't know who the real winner was. To make matters worse, Florida was not the only state where this situation occurred; it was just the state that got the most attention.

This is an opportunity for a group of engineers, technologists and business people to really make a difference in the world and in the United States by developing voting equipment that is cost effective and fault tolerant enough to make sure it is able to record and accept every vote cast.

It is apparent that the United States needs equipment and procedures that can eliminate the problems that have been reidentified during the most recent election season. I say reidentified, because these problems have been identified before.

Back in the early 1970s, my mother was a test subject for the now infamous "butterfly ballot." Some of the problems in this election were identified by her then. Even the punch card systems with chads had known problems. There are several patents in the U.S. patent system to eliminate pregnant and dangling chads.

To be successful, it needs to be a six-sigma solution. It needs to be sufficiently fault tolerant so that less than one error in a billion occurs. The six-sigma solution has been applied to many industrial situations from diapers to semiconductors to automobiles. There is no reason it can't be applied to the voting process.

To be successful and to meet all the stated objectives, it must also be cost effective enough for even the poorest voting districts. In the United States, the next trick is to have the solution installed in time for the presidential election in 2004. The challenge is now on the table.

— Niel Leon
Committee on Engineering
Entrepreneurship
leonn@asme.org

back to columns