The Effects of Task Partitioning and Computer Skill on Engagement and Performance with an Adaptive, Biocybernetic System

The present study examined the effects of computer skill and task partitioning on a tracking task using a biocybernetic, adaptive system. The tracking task was partitioned into horizontal and vertical axes, and the adaptive system allocated control of the axes between the human and the system in real time. Results showed that sharing a task with a computer with expert-level skill elicits performance comparable to performing the task manually. However, sharing a task with a computer with novice-level skill degrades performance compared to performing the task manually. Thus, although no clear advantage was found for being paired with an expert-level computer, there was a definite disadvantage to being paired with a novice-level computer. In addition, the present study also showed the potential improvements in system behavior by manipulating the automation switching criteria, namely increasing the range around the baseline EEG engagement mean.