Work-Related Fatalities Analysis Through Energy Source Recognition
Document Type
Conference Proceeding
Publication Date
1-1-2020
School
Construction and Design
Abstract
Identification of hazards is the first step in accidents prevention. As hazards can be caused by different energy sources, the awareness of all the energy sources is key to identifying potential hazards and creating a safe environment. Accidents result from the interaction of energy, equipment, or materials, and one or more people, and the potential hazards associated with such interaction can be identified based on the energy sources recognition. A lack of understanding of the presence and magnitude of an energy source often results in an accident. As a result, it is important to identify highly innovative and effective hazard recognition strategies such as implementing techniques to avoid future accidents. This study analyzes fatalities and catastrophes data inspected by federal or state OSHA in the past 5 years. Modern machine learning techniques are deployed to power this study: 1) text mining for hazard report extraction and 2) multidimensional visualization for geospatial analysis. The outcome can assist personnel involved in high-risk activities to identify and control the potential hazards unique to each activity and job. As a result, energy sources posing dangers would be effectively managed and eliminated.
Publication Title
Construction Research Congress 2020: Safety, Workforce, and Education - Selected Papers from the Construction Research Congress 2020
First Page
279
Last Page
288
Recommended Citation
Song, S.,
Awolusi, I.,
Jiang, Z.
(2020). Work-Related Fatalities Analysis Through Energy Source Recognition. Construction Research Congress 2020: Safety, Workforce, and Education - Selected Papers from the Construction Research Congress 2020, 279-288.
Available at: https://aquila.usm.edu/fac_pubs/19123