Using human-AI collaboration to monitor PPE donning, doffing

Editor's Note

This Australian study compares the accuracy of monitoring personal protective equipment (PPE) donning and doffing between an artificial intelligent (AI) machine collaborating with a remote buddy support system and an onsite buddy.

The researchers conducted a pilot simulation study with 30 procedural scenarios (15 donning and 15 doffing) and assessed 195 steps.

Overall accuracy of PPE monitoring with the human-AI system was 100%, and the overall accuracy with the onsite buddy was 99%.

The AI machine was able to perform autonomously, without any rectification from the remote buddy in 173 (89%) of 195 steps, and it identified 67.3% of all errors independently.

The findings provide preliminary evidence that a human-AI machine system may be able to substitute or enhance an onsite buddy performing monitoring of PPE donning and doffing, the researchers say.