Demeter – a Risk Mitigation Tool for Agriculture Workers

ABSTRACT

The agriculture industry lacks novel techniques for analyzing risks facing its workers. Although injuries are common in this field, existing datasets and tools are insufficient for risk assessment and mitigation for two primary reasons: they provide neither immediate nor long-term risk mitigation advice, and they do not account for hazards which fluctuate daily. The purpose of Demeter is to collect safety data about hazards on farms and produce risk analysis and mitigation reports. This application uses a combination of formula-based risk calculations and state-of-the-art graph neural networks (GNNs) to perform risk analysis and reduction. The formula-based risk calculations had a mean absolute error (MAE) of 0.2110, and the GNN had an accuracy of 94.9%, a precision of 0.3521, and a recall of 0.8333. Demeter has the potential to reduce the number of injuries and fatalities among agriculture workers by alerting them to risks present in their daily workflow and suggesting safety precautions.

Keywords:

• Deep Learning
• Graph Neural Network
• risk analysis
• agriculture worker safety

Acknowledgments

We would like to express our gratitude to Dennis Murphy, Ph.D and our MCRI partners for their subject matter expertise and support in data collection.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was supported [in part] by the intramural research program of the U.S. Department of Agriculture, the National Institute of Food and Agriculture [Phase I SBIR: Risk Management Tool to Increase Farm Safety, Contract No. 2022-40000-36950].