Modern submarines deep underwater arguably have better situational awareness than today’s drone pilots. Research has demonstrated that small unmanned aircraft are all but invisible from manned cockpits for collision avoidance purposes, and even manned aircraft are difficult to see from the ground with time left to give way. These are among the facts AOPA considered when commenting on landmark changes to Part 107.
AOPA Senior Director of Airspace, Air Traffic, and Aviation Security Rune Duke noted in the association’s formal responses to a pair of rulemaking initiatives proposed by the FAA that freedom to fly for all is among our top priorities. That includes the hundreds of thousands of manned aircraft pilots among our membership; the tens of thousands who fly small, unmanned aircraft as well; and thousands of members who are Part 107 remote pilots only, who joined after AOPA first extended a membership offer to those who fly only drones in February 2017. AOPA has since become one of the world’s largest organizations of unmanned aircraft pilots as well as more traditional aviators.
The FAA announced both rulemaking actions in January, inviting comment by April 15 on the ANPRM as well as on the NPRM, making clear the agency's view that any new expansion of allowed drone operations also depends on establishing rules and standards for remote identification and tracking of unmanned aircraft.
Ongoing research by Embry-Riddle Aeronautical University, Oklahoma State University, and Kansas State University, among others, has produced a growing body of evidence and data that make it clear “see and avoid” remains a huge challenge for human eyes, day or night, with or without strobes, even when manned aircraft pilots are told in advance where to look for the drones flown in a series of real-world trials. One study published in 2017 found that ground observers were able to reliably detect manned aircraft at a distance of about 1,600 feet or more, though participants did not perform well when estimating intercept time and distance.
“During 26 of the intercepts, participants overestimated the available duration to the aircraft-sUAS intercept; and, underestimated the duration during 11 of the aircraft-sUAS intercepts. The mean estimation error was determined to be 17 seconds,” the authors wrote. “Given the relatively short intercept time from when initial visual detection is made, any further delay incurred from overestimation may not leave the remote pilot adequate time to react to evade a collision threat.”
Embry-Riddle professor Ryan Wallace, among the authors of this and related research published so far, said human vision is even less likely to detect and correctly identify a small drone from the cockpit of an aircraft in flight. Flight crews were unable to reliably detect small drones until they came within about 500 feet of the sUAS in real-world trials, far too close for effective action to avoid collision.
“You don’t have time to react,” Wallace said in an interview. “Unfortunately, see and avoid is all we have.”
Further research, though inconclusive, suggests strongly that strobe lights do little if anything to make drones more visible to manned aircraft during daylight. The size of the unmanned aircraft and limits of human optical perception both factor in making drones effectively invisible to almost any manned aircraft, almost certainly so to fixed-wing pilots. Only a slow-moving helicopter stands a reasonable chance of seeing and avoiding a drone, and the high workload and unique demands of low-altitude operations also work against manned helicopter pilots, who have made no secret of the fact that drone pilots simply cannot expect their tiny aircraft to be seen.
Duke offered suggestions for improving the proposed new regulation for nighttime operations under 14 CFR Part 107.
“We believe additional lighting equipment to be necessary for the sUAS operator to maintain situational awareness. Human factors research is clear that detecting relative motion, such as determining an approaching manned aircraft, can be more challenging at night than during the day,” Duke wrote. “At night, even with the sUAS equipped with anti-collision lighting, we remain concerned whether a pilot in a manned aircraft will be able to recognize and avoid a collision with an sUAS. An sUAS is likely to disappear amongst city or country background lighting.”
Duke wrote on AOPA’s behalf that so-called “non-cooperative” aircraft, typically older or smaller airplanes not equipped with transponders or Automatic Dependent Surveillance-Broadcast Out capability, will continue to share the airspace for years, perhaps decades to come, and any system that increases their see-and-avoid burden is set up to fail. Instead, the FAA should leave that burden on remote pilots and drone makers, leveraging technology when possible but also establishing some degree of operational segregation, such as maintaining the current 400-foot altitude limit for drones, and the current 87-knot sUAS speed limit, since collision risk increases with velocity.
“Stand-off distances integrated into sUAS aircraft via voluntary technology, combined with operator education and common-sense regulations, will mitigate the risk of an accident as the number of sUAS operations increase,” Duke wrote. “We believe the FAA should be careful to not take a reactionary approach and curtail UAS operations or overregulate this emerging aviation industry. The operational and policy dictated reaction must be commensurate with the threat… We believe the FAA should enact performance-based rules to account for different technological solutions, like detect and avoid, that are shown to be equivalent or better than human performed see-and-avoid.”
As the FAA ponders allowing more advanced unmanned operations including flights beyond the pilot’s visual line of sight (expected to be the subject of a future proposed rule once a system of remote identification and tracking of drones is established), there is disturbing evidence that at least a few remote pilots are already flying beyond the current limits. Wallace was among the authors of a study published recently for which researchers collected drone telemetry data and determined that 5 percent of small unmanned aircraft flown over the course of the study period were flown so far from their launch point that it would be impossible for a ground observer to see them.
AOPA views education of all airspace users as critical to safe integration and noted that technology has many limits.
“We do not believe the see-and-avoid requirement for UAS operators can be accomplished solely on the reliance of general aviation aircraft having a transponder or ADS-B. From both the manned and unmanned perspective, there should be concern and caution regarding overreliance on cooperative systems for detecting nearby manned aircraft,” Duke wrote. “Not only will engine-driven general aviation aircraft not necessarily be equipped, but ultralights, gliders, birds, and other obstructions will occupy the airspace. It is important technological solutions are fielded that account for non-cooperative aircraft and objects, and the philosophy of cooperative deconfliction should be relegated to the minority of special cases.”
Current detect-and-avoid technologies being tested include drone-mounted microphones, cameras, and radar, and many experts believe a robust and reliable solution will involve some combination of these approaches, along with software-imposed limits on where drones can fly known as “geofencing,” which relies on GPS.
“This is a critical technology that has been integrated into flight control systems, navigation, geofencing, ADS-B, etc., such that it becomes a single point failure that could lead to disastrous results,” AOPA noted with regard to GPS. “The FAA should engage with UAS users and the government agencies that conduct GPS interference events to educate and determine any necessary mitigations.”
AOPA also urged the FAA to remain agnostic about which technology is used in a given drone, and focus requirements on performance-based standards that apply to all: “We do not believe that a single manufacturer’s limitations should negatively affect the entire industry. Regardless, the FAA must ensure restrictions are in place to ensure manned aircraft are adequately protected from a midair collision with a UAS.”
Congress repealed in October a law that had previously limited the FAA’s freedom to regulate noncommercial drone use, and the Special Rule for Model Aircraft established by the FAA under that former law was effectively scrapped in April. Recreational drone users are required to register their aircraft (if the aircraft weighs more than .55 pounds) and follow additional rules, including the 400-foot altitude limit that was previously a guideline for noncommercial drone flyers. Additional rules will eventually be put in place to mitigate risks posed by recreational operations, though the current rulemaking applies to Part 107 operations only.
The FAA envisions allowing commercial drone operations to venture farther afield in the future, and new products and technology are being developed almost daily. A Swiss rescue organization’s autonomous, helicopter-style drone is but one of dozens of examples. Scaling such beneficial uses up will require more than detect-and-avoid technology—it will require an air traffic management system similar to the current network of radar, air traffic control facilities, and air crews who coordinate thousands of flights each day.
NASA is leading an effort to develop a traffic management system for unmanned aircraft that is analogous to the systems currently used to assist in the safe separation of manned traffic. AOPA cautioned that this unmanned aircraft traffic management (UTM) paradigm must not require manned pilots to participate in an operational sense, though collaboration on system development should increase.
“We believe it would be premature to implement UTM without further collaborative work with manned aviation. The UTM system must be transparent to general aviation and ensure manned operations are not required to participate,” AOPA wrote. “Manned aviation already operates in an air traffic management system managed by FAA, military, and contract air traffic controllers. It would be confusing and onerous to impose a new system on general aviation aircraft that fly at low-altitude. There should be no expectation that manned pilots will need to review UTM information before flight.”
Many AOPA members are among those who enjoy flying a drone purely for fun, and should not be forced to participate, Duke added.
“Hobbyist sUAS operations will continue under visual line of sight operations and are not going to be at the volume of commercial operations, so we do not believe it should be necessary for recreational sUAS operators to need to participate except from an advisory or voluntary standpoint.”
While drones have proliferated in recent years and sales of these inexpensive aircraft continue at a brisk pace from big box stores and online retailers alike, documented collisions between manned and unmanned aircraft remain rare.
Notable exceptions include a U.S. Army Black Hawk helicopter colliding with a small quadcopter over New York in 2017, a case in which the untrained operator flew far beyond line of sight; a hot air balloon was struck by a drone over Idaho in 2018, also a case involving an inexperienced operator who lost sight of his aircraft. In these cases, and a handful of other reported collisions, there was no injury or loss of life, though the potential has been established by FAA-sponsored research that used computer models to simulate a range of possible collisions, some of which resulted in potentially catastrophic structural damage to the manned aircraft.
AOPA continues to support education and training, a methodical approach to expanding operational limits for drones that keeps safety squarely in focus, and avoidance of onerous requirements including monopolization of UTM by any one company that would make participation unnecessarily expensive for hobbyists and remote pilots.
“AOPA seeks to protect the freedom to fly for everyone who loves to fly, regardless of the aircraft they choose,” Duke wrote. “We are actively working to safely integrate UAS operations within the NAS by ensuring that these pilots have the appropriate level of aeronautical knowledge.”