Authors:

Arghya Sarkar

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

As automation becomes more prevalent across industries, achieving effective coordination between humans and autonomous systems is crucial. Human decisions to rely on automation are influenced by human cognitive states such as trust in the automation, self-confidence in task execution and workload. Furthermore, it has been observed that both under-reliance and over-reliance on automation pose risks. Consequently, there is a need to develop autonomous systems that can appropriately calibrate human cognitive states in human automation interaction contexts. To verify the relationship between workload and self confidence in human automation teaming (HAT) contexts, we built a search and rescue game in which the number of roles assigned to the human and automation varies. A probabilistic model of human self-confidence and workload is to be trained from the data by using a Markov decision process framework. Additionally, an optimal control policy is developed using a reward matrix to maximize HAT task performance and calibrate human cognitive states. The implications of this research lie in the improved understanding of workload and self-confidence in HAT scenarios. Future research should focus on applying these findings in practical settings and further refining the calibration algorithms.

Source:

Purdue University / 2023

Topics:

No topics listed

Co-authors:

Arghya Sarkar