Analytical models and big data may be used to evaluate the relationship between key leading indicators of health and safety (H&S) and their downstream lagging indicators. Lagging indicators are being considered that include: 1) Eye injuries; 2) Hand injuries; 3) Sprains and strains of the ankle, knee, shoulder or back; and 4) Fractures and amputations. Using data provided by our industry partners, we employ machine learning (ML) methods to examine underlying relationships. This work includes two phases: Phase 1 involves data collection and discovery of leading indicators using both exploratory and confirmatory statistical approaches. Phase 2 involves developing a proof of concept predictive model for each lagging indicator and assessing a variety of ML approaches. Classifier accuracy will be evaluated against subject expert baselines as well as H&S outcomes. The leading indicators and predictive technologies developed in this work will serve as a foundation to augment existing training programs and control hierarchies and develop proactive performance dashboards for next-generation H&S management systems.  

Seed funding has been provided by the new School of Mining & Mineral Resources to begin this project.