Is this project an undergraduate, graduate, or faculty project?
Undergraduate
Project Type
individual
Campus
Daytona Beach
Authors' Class Standing
Katherine Hoffsetz, Junior
Lead Presenter's Name
Katherine Hoffsetz
Lead Presenter's College
DB College of Arts and Sciences
Faculty Mentor Name
Dr. Mihhail Berezovski
Abstract
Spatial disorientation (SD) is a critical factor contributing to accidents in general aviation (GA), posing significant challenges to flight safety. This study conducts a comprehensive data analysis of general aviation crash reports to investigate the patterns and underlying factors associated with spatial disorientation. The dataset, derived from official reports from the National Transportation and Safety Board, has over 800 cases to study. Utilizing advanced statistical methods and machine learning algorithms, we analyze the textual data from official accident reports to extract and categorize instances of spatial disorientation. The methodology involves natural language processing (NLP) techniques to systematically review and classify the narratives and findings within the reports, focusing on keywords and phrases indicative of SD. Additionally, we correlate these instances with various factors such as weather conditions, flight phase, pilot experience, and aircraft type to uncover significant trends and risk factors. This study contributes to the ongoing efforts to enhance GA flight safety by providing a data-driven understanding of spatial disorientation accidents.
Did this research project receive funding support (Spark, SURF, Research Abroad, Student Internal Grants, Collaborative, Climbing, or Ignite Grants) from the Office of Undergraduate Research?
No
Investigating Spatial Disorientation Related Aviation Mishaps Utilizing Machine Learning Keyword Extraction Models
Spatial disorientation (SD) is a critical factor contributing to accidents in general aviation (GA), posing significant challenges to flight safety. This study conducts a comprehensive data analysis of general aviation crash reports to investigate the patterns and underlying factors associated with spatial disorientation. The dataset, derived from official reports from the National Transportation and Safety Board, has over 800 cases to study. Utilizing advanced statistical methods and machine learning algorithms, we analyze the textual data from official accident reports to extract and categorize instances of spatial disorientation. The methodology involves natural language processing (NLP) techniques to systematically review and classify the narratives and findings within the reports, focusing on keywords and phrases indicative of SD. Additionally, we correlate these instances with various factors such as weather conditions, flight phase, pilot experience, and aircraft type to uncover significant trends and risk factors. This study contributes to the ongoing efforts to enhance GA flight safety by providing a data-driven understanding of spatial disorientation accidents.