An Evaluation of a Personalized Automated System Designed to Detect Hazardous Noise Exposure Among Farmers

Occupational hearing loss (OHL) among farmers is higher than in most other occupations. The HearSafe study was a hearing loss intervention study that aimed to increase the use of hearing protection devices (HPD) among farmers by providing personalized hazardous noise information through the use of the HearSafe system. The HearSafe system consisted of a wearable noise sensor that recorded loud sound “events” and a website with a programmed computer algorithm. The base algorithm on the website was programmed to group events into “tasks” if events occurred within two minutes of each other and group tasks together if tasks were within an average distance of 30 ft of each other. The ability of the HearSafe system to increase the use of HPD among farmers is dependent upon the accuracy by which the computer algorithm groups tasks. Ensuring high task grouping accuracy will improve the information provided to the farmer and thereby increase the potential for the HearSafe system to increase the use of HPD among farmers.

The aim of this study was to increase task grouping accuracy of the HearSafe system. Multiple alternative parameters were programmed into the algorithm and evaluated for task grouping accuracy. Results were compared to the task grouping accuracy of the base algorithm.

The algorithm that produced the greatest increase in task grouping accuracy was the model that combined tasks that had consecutive events within 30 ft of each other with a programmed time interval of 10 min between events and an average distance of 30 ft between tasks. The results of this study improved the ability of the HearSafe system to automatically group tasks by increasing accuracy from 57% to 87%. An improved ability to accurately group tasks will increase the potential for the HearSafe system to effectively increase the use of HPD and, therefore, decrease the prevalence of OHL among farmers.