Airspace regulators around the world are concerned about incidents between drones and manned aircraft – and a dive into the data found in the recently released FAA Administrator’s Fact Book for 2019 (Fact Book) shows why. Is the data accurate enough to use?
Accident Data and the U.S. Safety Record
The National Airspace (NAS) in the U.S. is some of the busiest in the world: but it is also one of the safest. While development of new regulations for unmanned systems may at times have been slower than the drone industry would have liked, the FAA is defending a hard-won and impressive record for safety. Accident rates are stunningly low. According to the data in the Fact Book, there were fewer than 35 accidents per year in commercial airlines from 2006 – 2015; most years there were absolutely no fatalities.
The National Transportation Safety Board figures, which display 2016 and 2017, also show no fatalities at all for commercial airlines during those years. Compared to other modes of transport – highway, railway and marine – flying on a commercial airline in the U.S. is by far the safest way to travel.
It isn’t just numbers of fatalities that are low (although travelers might argue that’s the most significant data set). Accident rates during the years 2006 – 2015 are correspondingly low. Accident rates per 100,000 departures range from 0.25 to 0.35 (rounding to two decimal points is mine) each year – a rate unthinkable on the highway.
Accidents vs. Incidents: The Data on Drones
Accidents – which remain as stated above, thankfully, extremely rare – are not the same as “incidents.” Tables of Air Traffic Incident data are designed to show trends in risk. As such, they are defined in terms of how close one aircraft came to another, or whether a pilot was forced to deviate from the plan and take “evasive action” to avoid a collision.
It’s in rates of “incidents” that the data about drones is raising safety concerns all over the world. Specifically, in one category of incident: Near Mid-Air Collisions, or NMACs. As defined in the Fact Book:
Near Mid Air Collision (NMAC): when an aircraft flies within 500 feet of another aircraft, or a pilot or flight crew member reports a collision hazard between two or more aircraft.
NMACs stayed constant over 2014 and 2015, at 143 and 145 – that’s total, for the year – respectively. But in 2016 and 2017, those incident rates more than doubled, to 302 and 385 per year, respectively. The blame for the rise, says the FAA, can be laid entirely on drones. “Pilot-reported NMACs with Unmanned Aircraft Systems (UAS) now account for more than half of all reported events,” says the FAA.
The data tables are disturbing; and they are influencing regulations and public opinion around the world. According to the numbers, the risk – not the number of accidents – is rising sharply. In including drones in the data as equivalent with manned aircraft, however, the numbers may no longer be accurately reflecting real risk to air travelers.
The Breakdown in Data: Reporting
One of the major problems with drone incident data is in reporting. While pilots can easily, obviously, and accurately report another airplane at 500 feet; they cannot easily and accurately report a drone at 500 feet. Well-publicized reports of “near misses” with drones have later been found to be balloons, plastic bags, birds, or other objects.
Major drone manufacturer DJI and other stakeholders have struggled to make the point to regulators around the world. When it comes to clearly identifying an object as small as a modern commercial drone (not to mention an even smaller recreational drones) from a distance of more than one and half U.S. football fields, or the length of a skyscraper more than 40 stories tall – most humans just can’t be that reliable.
The Real Risk in Collision: Still Unknown
Aside from a critical issue of accuracy in numbers, a major problem with drone incident data is that it doesn’t necessarily equate to risk in the same way that the NMAC standard as applied to other manned aircraft does. If two manned aircraft come in close proximity to one another, there is a real risk of collision – and a real risk that the collision will significantly damage both aircraft, placing passengers in danger. In that case, an increase in near misses has a direct correlation to increased risk of accident.
The same isn’t necessarily true for drones. There have been studies involving computer simulations, mathematical assumptions, and even crashing old drones into even older airplane parts. In the past five years, such studies have abounded. They have come to clear and well-publicized conclusions: but those definite conclusions that have ranged from “catastophic risk” to “absolutely negligible danger.”
The fact remains that we still don’t know what would happen if a drone hit a manned aircraft in exactly the right place or exactly the right way. A very few real collisions between manned aircraft and a drone (such as the blackhawk helicopter incident) have actually occurred. They have not caused passenger injury or major damage to the aircraft.
The Search for a New Standard
That’s not a claim that drones pose no danger or risk. It’s just a statement that as of yet, there seems insufficient evidence to reach a conclusion upon which to base long-reaching regulations. And, as the life-saving applications of dronesevolve and are more commonly implemented, the risk of limiting those operations must be also be weighed against the risk to manned aircraft. According to a conservative record of lives saved directly due to drones, drones saved more than 65 people last year.
As new FAA Administrator Steve Dickson steps into office tasked with dealing with the recent Boeing scandal and the daunting job of modernizing the air traffic system, he may also look for data from which to base important decisions about protecting the U.S. airspace safety record. Including drone incident data into the regular tables of manned aircraft incident data may not provide enough information for the purpose.