Learning curves: Literature review and further research

Abstract

Learning curves have been proven to be effective tools for monitoring the performance of workers exposed to a new task, describing their development as task repetition takes place. These curves were originally proposed by Wright in 1936, and have been used since then to evaluate the time demanded to complete production runs, to estimate production cost reduction as learning takes place, and to assign workers to tasks based on their performance profile. This paper presents the state-of-the-art in the literature on learning curves, covering the existing models, their limitations, and reported applications. Further research on the subject is proposed.