A pooling design is an algorithm to intelligently classify items by testing them in groups or pools rather than individually. The result from the pools is usually binary — either positive or negative. A negative result can imply that all the items tested in that pool were failures, if the tests are error-free. A positive result implies that at least one of the items in that pool was a success, under the same error-free assumption.
The central idea originated from the practice of blood testing cadets for syphilis during World War II [1]. It has applications in communication systems, molecular biology and general fault testing areas.