Every day, work-related injury records are generated. The National Institute for Occupational Safety
and Health (NIOSH), part of the Centers for Disease Control and Prevention (CDC), recently announced
the start of an external crowdsourcing competitionexternal icon?, exploring the use of artificial
intelligence (AI) to automate data processing in occupational safety and health surveillance systems.
Currently, whenever an employee is injured at work, an explanation of how the injury occurred is
recorded by a person using free-text narratives, or rather, free form. For decades, humans have then
read these injury narratives and assigned codes to classify the injuries, often large volumes of
information, which has resulted in time, cost and the risk of human error influencing occupational
safety and health data.
Through an interagency agreement with the National Aeronautics and Space Administration (NASA)'s
Tournament Lab, NIOSH is working with vendor TopCoder to host this online competition, asking
programmers to compete in the development of an algorithm that will best employ the use of AI in
automatically reading injury records and classifying them according to the Occupational Injury
and Illness Classification System (OIICS).