Drones are coming to airspace that private and commercial aircraft presently use.  What are the safety, risk, and insurance problems this raises?

I’ve been talking with pilots lately about drones.  The pilots in my social circle fly small airplanes; Cessnas and Pipers and Bonanzas.  Except for the occasions when I get to say, “Goodbye” to the Captain while I’m deplaning a commercial flight, I don’t have a lot of opportunity to talk with pilots who fly the really heavy stuff.  But the recreational pilots — the ones who will, of a Sunday morning, fly from my little home airport in Somerville, NJ to, say, Portland, ME to lunch on lobster salad — are universally worried about the commercial use of what are known as unmanned aircraft systems (or UAS) in “our” airspace.  As one such pilot recently told me, “I have enough on my mind when I’m flying that I shouldn’t have to worry about a small metal object being flown by someone on the ground who is talking to no one, possibly being regulated by no one, oblivious to my presence or my intentions, and that will be invisible to me right up until the instant when it collides with my airplane.”

The commercial use of radio-controlled UAS is well on the way, and the perceived risks of this new and burgeoning activity have a lot of general aviation (or GA) pilots wondering if there’s anyone in charge who has yet thought this thing through.

In December 2013, the Federal Aviation Administration selected six “test sites” for the study of the commercial use of UAS.  The selection was part of a Congressional mandate under the FAA Modernization and Reform Act of 2012, which required the FAA to establish a test site program to “integrate UAS into the National Airspace System by September 2015.”  The six test sites are the University of Alaska; the State of Nevada; Griffis International Airport in Rome, NY; the North Dakota Department of Commerce; and Texas A&M University, Corpus Christi.  The test sites are supposed to come “on line” sequentially and to operate until February 2017.


The commercial use of drones is on the way.


 Several days ago, the FAA issued what is called a Certificate of Waiver or Authorization to the North Dakota Department of Commerce, the first of the test sites to begin operations — or to “stand up,” in FAA jargon.  Test flights of UAS are scheduled to begin this week over the North Dakota State University Carrington Research Center and Sully’s Hill National Game Preserve near Devils Lake, ND.  The goal of the North Dakota initial operations is to show that UAS can check soil quality and crop status.  Although it is not hard to imagine a great many commercial uses for UAS, such as inspection of buildings, power lines, and facilities located in remote or forbidding terrain; surveillance by state and local police forces; border patrol; and aerial photography, to name just a few, agricultural uses are currently a key proposed application of UAS.

A goal of all of the test site operations, according to the FAA, is to find “solutions for ‘detect and avoid.’”  In other words, to find a way for UAS to detect objects in flight and to avoid flying into them.

Here are some of the reasons GA pilots are spooked about sharing airspace with UAS.  To begin with, there are approximately 6,000 airports in the continental United States, the overwhelming majority of them small private or municipal fields that service recreational pilots, flight schools, and local charter operations.  Pilots who use most of these airports cannot talk to air traffic controllers in control towers because these airports, for the most part, don’t have control towers.  Instead, they have what are called common traffic frequencies that are printed in official FAA publications and on which pilots voluntarily communicate with one another over their radios.  UAS, of course, being unmanned, are inherently uncommunicative.

Also, GA aircraft often fly at low altitudes where commercial UAS are expected to do much of their work.  Obviously, when planes are in the act of taking off and landing, they are, in aviation parlance, low and slow.  The traffic pattern altitude for GA aircraft landing at an airport is typically between 1,000 to 1,200 feet above ground level.  What is more, a great many small local airports, which service the GA population exclusively, are located in rural areas, often surrounded by farmland — precisely the areas that UAS manufacturers are targeting as markets for their aircraft.  The risk of unhappy interaction between UAS and GA aircraft is real and obvious.  It is small wonder that this has GA pilots worried.

In fact, there is a small uncontrolled GA airport in Carrington, ND, where the UAS test flights are to begin this week.  There is another small uncontrolled airport in Devils Lake, ND, the second site of the imminent North Dakota test flights.  When the tests at Griffis International begin, the UAS will be flying out of an airport that is, itself, surrounded by other airports.  Syracuse airport is 30 nautical miles (about 35 statute miles) to the west.  A tiny uncontrolled airport called “Becks Grove” is 9 NM away.  Another small uncontrolled airport, Frankfort Highland, is 16 NM from Griffis and little Hamilton airport is 23 NM away.

It may come as a surprise to non-pilots, but it is perfectly legal for pilots of general aviation aircraft to fly their planes in the airspace below 18,000 that blankets approximately 95% of the United States without ever talking to an air traffic controller, or to other pilots, for that matter.  The most important Federal Aviation Regulation to all GA pilots in flight is the requirement to “see and avoid” other aircraft.  There are specific protocols and conventions that each pilot of accidentally converging aircraft are supposed to follow so that the pilots of each plane will not be surprised by another pilot’s maneuvers and so that aircraft can avoid colliding in mid-air.  Thus, the UAS program to find “solutions for ‘detect and avoid’” are unnerving.  The question that comes to mind is: Isn’t there a detect and avoid system already in place, before the test flights begin?  Other questions flow from this one.

Will the FAA require all UAS to have the capability to detect and avoid other aircraft and how will that capability interact with what a pilot might do if he or she detects an oncoming UAS first?  What happens if detect and avoid systems fail or, worse, if they are hacked and maliciously disabled?  How will detect and avoid be enforced and will the FAA have the resources to do so?


The Air Force’s experience flying UAS supports the concerns GA pilots are feeling.


The anxiety among pilots concerning UAS is intuitive and not based upon experience because UAS have not yet been integrated into the National Airspace and the commercial use of these aircraft outside the National Airspace is nascent.  In other words, we don’t yet have experience with the use of UAS for commercial purposes.  The Air Force, however, does have experience flying UAS in military airspace and its experience supports the concerns GA pilots are feeling. (See the Air Force’s Accident Report for Fiscal 2014 here.)

Not surprisingly, the accident rate for Air Force operated UAS increased dramatically from 2004 to 2013 when the Air Force substantially increased the number of such flights.  What is more, the rate of mishaps involving UAS was much higher than the rate involving manned aircraft over that same period.  GA pilots, then, have reason for concern about the risk to themselves and others if and when large-scale commercial UAS operations  begin in the National Airspace.

Whenever I think about risk, I think about insurance.

So what kinds of insurance are UAS manufacturers and operators going to need?  If and when farmers start using UAS to help manage their crops, whether they operate the UAS themselves or contract the job out to specialized operators, they, too, are going to need insurance against accidents involving UAS and GA aircraft.  Risks include the obvious: product liability, bodily injury (including death) and property damage; as well as the not-so-obvious, such as invasion of privacy.  UAS have also become increasingly sophisticated in the collection, processing, and storage of vast amounts of data, a fact that makes them both enormously useful and subject to data breach and cyberattack.

The standard Commercial General Liability policy that most businesses purchase covers bodily injury and property damage caused by an “occurrence,” which it defines as “an accident, including continuous or repeated exposure to the same generally harmful conditions.”  Essentially all such policies, however, have exclusions for damage caused by the operation of aircraft.  UAS manufacturers and operators, like airlines and aircraft manufacturers, are going to need to be covered by specialized liability policies.

There are, already, insurance companies offering coverage for manufacturers and operators of UAS.  And policyholders can purchase up to $100,000,000 in coverage.  But those policies currently cover risks that are vastly different in nature and scope from the risks associated with extensive commercial UAS operations in the National Airspace.  If and when such operations begin, insurance underwriters will search in vain, at least for the first 5-10 years, for any kind of experience that will help them to assess the risks accurately and to price the coverage meaningfully.  Companies such as Underwriters at Lloyd’s, which has historically provided coverage for risks that other carriers would not (think Betty Grable’s legs), will likely get into the market to cover commercial UAS operations and may begin by writing coverage that is unique to each manufacturer’s or operator’s business (known as “manuscript” policies).

The kinds of coverage UAS manufacturers and operators will need include first-party property, for damage to the UAS, itself; third-party liability, for injury or property damage to others; cyber insurance for data breach, invasion of privacy, and disclosure of private information; and possibly business interruption and contingent business interruption.

Insurance, however, is nothing more than protection against the occurrence of a catastrophe.  It comes into play only after the fact.  It is the unknown — and possibly unknowable — risk of accidentally encountering UAS in the air that has pilots fretting about sharing airspace with remote-controlled metal objects.  And the existence of insurance to pay for the consequences of such an encounter afterwards is small comfort.

Here’s hoping the FAA knows what it’s doing.