Electronic Thesis/Dissertation
 

A Machine Learning Approach to Unmanned Aircraft System Collision Risk Assessment

Open Access

This project examines the potential hazards of collisions brought by Unmanned Aircraft Systems (UAS) intruding into controlled airspace. This study helps to fill the gap in airspace violation and collision risk assessment of all types of UAS as opposed to strictly small UAS. The insights from this study will be useful in improving airspace violation and incident risk assessment and management of UAS into controlled airspace. The resulting predictive model(s) can be used by decision-making authorities to evaluate current policies to prevent UAS collision and or/or incident to Manned Aircraft Systems (MAS) within controlled air space. Decision-making authorities can use the predictive model itself or use the insights from this study about the factors that would result to higher risk of collision among UAS. Current aviation detection systems are inadequate in effectively identifying UAS and therefore leave the detection of UAS to pilots or associated crew. Current studies have shown that the accuracy levels of pilots and associated crew to perform this task is below 50% which is well below any acceptable level of risk when there are human lives involved. Machine learning classification models were be utilized that are able to effectively create risk probability zones for violations and/or incidents allowing policy makers to better establish flight paths, take off times, and other import attributes of aircraft operations. This research has introduced several highly accurate models for predicting the probability of an incident or airspace violation based on known attributes as determined by statistical testing. Random Forest and Gradient Boosting algorithms performed with accuracies over 90% and the results show that these models will prove extremely useful in the establishment of risk probabilities and policy management.

Author Language Keyword Date created Type of Work License
  • All rights reserved
Rights statement GW Unit Degree Advisor Committee Member(s) Persistent URL

Notice to Authors

If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.

Thumbnail Title Date Uploaded Visibility Actions
Preview of Stracener_gwu_0075A_16130.pdf Stracener_gwu_0075A_16130.pdf 2022-10-04 Open Access