Predicting performance times for long cycle time tasks

TLDR

Long cycle time tasks comprise non‑repetitive sub‑tasks averaging about 1½ minutes, and performance degrades when a sub‑task reappears after a full cycle, so learning is predicted from the learning of individual sub‑tasks. The study proposes and tests a method to predict learning‑curve parameters of sub‑tasks and to model execution times for long cycle tasks. The method estimates forgetting as a function of the learning constant and interruption length and uses these estimates to predict sub‑task learning‑curve parameters. The resulting model accurately predicts execution times for long cycle tasks.

Abstract

A long cycle time task is assumed to consist of a series of non-repetitive unique sub-tasks whose standard times average at about 1 ½ minutes. 'Forgetting' is therefore a consequence of a specific sub-task reappearing in the next cycle after a whole cycle time of other activities is completed. Learning behavior of long cycle tasks is therefore predicted on the learning of its constituent sub-tasks. A method for predicting the learning curve parameters for the sub-tasks (the learning constant, and execution time of the first repetition) are proposed and tested. The extent of 'forgetting' is empirically determined as a function of the learning constant and interruption length. Finally, a model is developed for predicting execution times for long cycle tasks.