To Err is human: Building an automatic error generator system

Abstract

This paper introduces Human Error Generator (HEG), an erroneous human performance simulator which considers that the errors emerge from human-computer interaction. HEG, in its current version, has two implemented modules: one that designs the interface in which the simulated user performs a task; and an automatic human error generator module. The system is linked to the cognitive architecture ACT-R and is able to simulate the following errors during the user performance: perceptual confusion, omission, inversion, repetition and intentionality reduction. Furthermore, HEG generates a report with the error-affected human performance showing the information related to the cognitive process involved in the simulated task as output. This paper aims to contribute to the field of human-computer interaction regarding the development of systems that anticipate human error.