A 25-page Technical Report released by JVM, Product Development Group Manager. One of the early pioneers in optical mark and character reading, Jacob Rabinow, a Russian immigrant, is credited with the axiom, "It's not a question if paper feeding mechanisms will jam, mis-feed, or double-sheet, but if you can live with the incidence. Truer words were never uttered, when it comes to picking, feeding, and stacking paper documents at high speeds (or even low speeds, for that matter!). Our highest speed scanners, the W600 model line, scanned 8.5" x 11" sheets at 36,000/hour, and we ran MRC lab tests on prototype feeding mechanisms at up to 60,000 sheets/hour. A feed jam at this speed can cause serious damage to the offending sheet, and often several others, if the driving mechanism is not stopped quickly. Low humidity factors and/or foreign objects stuck the sheet (staples, candy, sweat, etc.) can cause two sheets to be picked together. Obviously, this condition must be rigorously detected to avoid a major scoring error or a missing record, or the front-side scores of student A's sheet to be erroneously combined with the rear-side of student B's sheet if these two sheets are stuck tightly together and pass through the dual-side scanning head on a single pass. Thus, the MRC engineering staff gave a lot of attention to multiple-sheet detection systems, devising several different methodologies, based upon mechanical scheme for thickness detection and/or optical opacity measurements (e. g., the difference between light transmission through one sheet versus two sheets). This report covers an optical solution. http://inventors.about.com/gi/dynamic/offsite.htm?site=http://web.mit.edu/invent/iow/rabinow.html