How AI Can Help Learn Lessons from Incident Reporting Systems

Both public agencies and private industry are learning from prior incidents using incident reporting systems to improve safety performance and reduce accidents. Incident reports are collected on anomalies of varying magnitudes—usually heavily weighted towards smaller incidents since they are more common than large ones. These reporting systems that collect information on anomalies are widespread across transportation, manufacturing, energy, and healthcare and are regarded as a foundational quality and safety improvement activity to which organizations are devoting considerable resources. For example, for decades, airlines and the Federal Aviation Administration (FAA) have placed a significant emphasis on voluntary data-gathering programs that enable airlines and government regulatory agencies to spot and correct problems before they lead to accidents. One such data collection system, the Aviation Safety Reporting System (ASRS) administered by the National Aeronautics and Space Administration (NASA), first established in 1976, receives more than 50,000 voluntary reports of safety incidents and situations each year from pilots, air traffic controllers, cabin crew, dispatchers, and maintenance technicians. Analyzing data from incident reporting systems can help an organization identify emerging risks, and this approach has been advocated for decades to prevent larger failures, but trends are difficult to find when there are more than 50,000 reports per year. This paper explores how AI tools can be used to find important lessons in incident reporting systems from near-miss events.