Drone spotting

The small, self-stabilizing and semi-autonomous aircraft commonly known as drones have become a popular consumer product in recent years, giving millions of newcomers access to aviation and airspace never before possible without considerable effort and training. Drones now clearly outnumber more traditional radio-controlled model aircraft, and manned aircraft as well. Despite the efforts of the FAA, AOPA, and other safety advocates, social media and video includes many examples of drones apparently being flown irresponsibly. One does not have to look hard to find a video of a drone climbing through clouds.

Sighting reports collected from flight crews and published by the FAA have often been cited by the press in the context of covering drones, while critics have noted that the data is often inconsistent, and missing key pieces of information that could distinguish a legal drone flight from one that is not. The Unmanned Aircraft Safety Team, a government-industry collaboration formed to study and promote safe operation of unmanned aircraft, took a deep dive into the available data and found that the same limitations that often apply to other types of near midair collision (NMAC) reporting, including challenges with estimating key variables like proximity, also apply to drone sightings, and that the resulting gaps in information frustrate any attempt to reach meaningful conclusions.

“We’re trying to get away from this perception that each sighting represents the same level of risk to the National Airspace System,” said Justin Barkowski, AOPA director of regulatory affairs and a participant in the UAST effort to analyze and report on the subject. “Many of these sightings could represent perfectly legal drone flights.”

The UAST analyzed 3,417 sighting reports published by the FAA from August 2015 through March 2017, taking a methodical approach to assess the risk. The group attempted to identify and define key parameters, such as the source of the report (flight crews, or other observers), the drone’s altitude, proximity to manned aircraft, whether evasive action was taken, and proximity to an airport, among others.

“While we believe the working group’s methodology helped provide valuable insights, ultimately the data set is too inconsistent and unstandardized to extract concrete conclusions,” the report states.

Among relatively few conclusions and calculations that could be reached, the report’s authors found 16 percent of the unmanned aircraft spotted were described by observers as being 500 feet or less from a manned aircraft, and just over 3 percent of those resulted in the manned aircraft changing course or altitude. In 70 percent of the sightings reported, the drone appeared to be above 400 feet. In most cases, 68 percent, the proximity of the drone to a manned aircraft could not be determined.

The finding that a relatively small percentage of the reports appear to reflect significant risk of collision with manned aircraft is consistent with other analyses, including studies of the same data by the Academy of Model Aeronautics and researchers at the Center for the Study of the Drone at Bard College. “There is general consensus that some of the sightings are potentially high risk and need to be mitigated, but the majority of sightings are not necessarily high risk,” the UAST report concluded.

The UAST made a few recommendations to improve the data collection, and inform future analysis. Potential measures to achieve that include defining what specific information should be collected from observers, establishing consistent standards for reporting and collecting that information (for example, should altitude be reported in reference to the ground, or sea level), and defining what criteria can be used to classify risk, “as opposed to creating a perception that each sighting represents a dangerous occurrence.”

The NTSB in December completed its investigation of the September midair collision between a DJI Phantom 4 and a U.S. Army Black Hawk helicopter, the first confirmed collision in the United States involving manned and unmanned aircraft. That investigation revealed gaps in the knowledge of the drone pilot (most notably, a lack of knowledge of the risk he posed to others by flying his drone far beyond line of sight). A 14-month effort to use computer modeling to assess the risk posed by midair collisions between drones and fixed-wing, manned aircraft determined that the structural damage that a drone can inflict on a manned aircraft is potentially more serious than some thought, though establishing the likelihood of such collisions was beyond the scope of that research.

Ongoing work to establish requirements for electronic drone tracking and remote identification promise to offer part of the solution, potentially giving law enforcement, the FAA, and manned aircraft pilots a tool to detect drones that can be very difficult to see with the naked eye, either from the ground or from a moving aircraft, and giving the FAA and safety advocates an objective source of reliable data on drone pilot misbehavior such as incursions into controlled airspace. Remote identification is also expected to promote accountability for drone operators.

The UAST report also recommends considering expanded efforts to educate the manned aviation community about drones, including visual identification of particular systems. Many sighting reports published to date indicate that the pilot reporting the sighting was unfamiliar with drones and what they look like, and the safety team noted that improving the quality of reports with more definitive information will be important as the number of drones increases. Improving the education of unmanned aircraft pilots is also important, the committee noted in its recommendations, suggesting that the FAA could leverage data collected from registered drone users to push information out to drone owners and pilots on a regular basis.

“There was a lot of really good discussion on educating both unmanned and manned communities in this area,” Barkowski said.