Many consider aircraft reliability a primary deterrent to the further integration of unmanned aircraft into civil airspace. Discussions of unmanned aircraft reliability often include comparisons of accident rates of medium and large unmanned aircraft (like the MQ-1 Predator and MQ-9 Reaper) to those of manned aircraft. These comparisons, however, often consider a recent period only but do not consider platform maturity. While this approach is valid—and worthy of discussion—it does not present a complete picture of the unmanned aircraft safety record. This paper employs a chi-square goodness-of-fit approach to compare the progression of the U.S. Air Force MQ-9 mishap rate to that of six manned aircraft flown by the same service. The analysis found that the MQ-9 did not have a significantly different Class A mishap rate progression from the comparison aircraft. The only exception was the F-16, which had a considerably higher rate. In addition, to better understand mishap causation, a comparison of manned to unmanned aircraft mishap causal factors was conducted, primarily relying on journal articles that use the Human Factors and Classification System (HFACS) taxonomy. The analysis found that, like manned aircraft, crew errors contributed to a significant portion of unmanned aircraft accidents, and further, that skill-based errors were often the type of the crew error. The studies also indicated that increasing autonomous control, Human Machine Interface (HMI) design, and crew training will continue to play a significant role in unmanned aircraft mishap causation. Two case studies were also analyzed to highlight differences in manned and unmanned aircraft accident causation.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.