Date of Award

Spring 2023

Access Type

Dissertation - Open Access

Degree Name

Doctor of Philosophy in Aviation

Department

College of Aviation

Committee Chair

David A. Esser, Ph.D.

First Committee Member

Jane Pan, Ph.D.

Second Committee Member

Mike F. O'Toole, Ph.D.

Third Committee Member

Kwok W. Chan, Ph.D.

College Dean

Alan J. Stolzer, Ph.D.

Abstract

Loss of control in flight (LOC-I) is one of modern aviation’s three most prominent fatal accidents. In the United States, air accidents are mandatorily reported to and investigated by the National Transportation Safety Board (NTSB). Established in 1976, the Air Safety Reporting System (ASRS) is a voluntary safety reporting (VSR) system administered by the National Aeronautics and Space Administration (NASA). Over 1.7 million ASRS reports have been processed to date. While the NTSB system handles LOC-I accidents, less severe incidents may have been reported voluntarily through the ASRS.

Safety reporting has been deemed the most valuable activity and the centerpiece of safety data collection for safety management systems (SMS). Both mandatory and voluntary safety reports (VSRs) are essential sources of SMS for safety assurance and risk management. Based on the age-old Heinrich’s common cause hypothesis, mitigating hazards identified in low-severity safety reports, such as voluntary safety reporting (VSR) programs, would prevent more severe events such as fatal accidents.

This mixed methods study aims to determine whether normalized rates of LOC-I hazards identified by NASA, named Belcastro LOC-I Hazards, differ collectively or v individually across mandatory and voluntary safety reports in the United States, represented by NTSB and ASRS reports. The quantitative part dominates this study. LOC-I safety reports were obtained from searches performed on already classified cases by the administrators of the databases, and by augmented search based on the LOC-I precursors keyword search used by Belcastro et al. (2017). A total of 12,432 safety reports from 2004 to 2020 were analyzed.

The research results suggested that the Belcastro LOC-I Hazard rates were statistically different at the multivariate level across the four safety report groups for both commercial and general aviation. Out of the eight Belcastro LOC-I Hazards, five in general aviation and seven in commercial aviation displayed univariate differences. A cursory review of the narratives of the reports also suggested that the textual reports related to the Belcastro LOC-I Hazards were contextually different across the groups. These findings provided insights: firstly, ASRS was a credible source in identifying some, but not all, hazards leading to LOC-I accidents; secondly, the augmented search would enrich intelligence gained from the ASRS database for some LOC-I hazards; and, thirdly, the validity of Heinrich’s common cause hypothesis was not generally supported.

While the NTSB system and investigations are more formalized, the research results suggested that ASRS safety reports are still effective in identifying some Belcastro LOC-I Hazards. This point is especially relevant in situations when accident data is limited. This research pointed to the need for a targeted approach, rather than one-sizefits- all, when using safety reporting databases. Before interrogating the data, practitioners should understand the precursors of the hazard to be analyzed, and the strengths and weaknesses of the associated safety reporting system. This awareness will enable safety vi professionals to calibrate, interpret, and supplement the data appropriately, resulting in more effective safety mitigations.

Download

Share

COinS