Top 5 Reasons Small Drones Crash More Often Than Manned Aircraft

General aviation remains one of the safest forms of transportation with a fatal accident rate per 100,000 flight hours of only 1.05 in the United States in 2013. The non-fatal accident rate per 100,000 flight hours was 5.85 during the same period. It isn't easy to make an apples to apples comparison with drone accident rates because, for the most part, small drone accidents happen often and most go undocumented.

2013 U.S. Aviation Statistics

Our estimated accident rate (non-fatal) for even the most experienced commercial drone operation would likely be on the order of approximately 1 accident every 250 flight hours for a rate per 100,000 flight hours of 400. There are many reasons for the wide gap compared to manned aircraft accidents but this article will focus only on what we see as the top 5.

#5 - Multirotor drones have very limited flight test data to learn from.
     
Today's small drones, for all intents and purposes, are very similar to RC aircraft. The first RC gliders, for example, have been around since the mid 30's. The first successful RC helicopter flight took place in 1969 by German engineer Dieter Schluter. Multirotor's however are a very recent development with the earliest versions only taking off within the last 8 years or so. An early pioneer in RC helicopters named Don Chapman actually described the basic multirotor concept to me in his Dayton Ohio basement way back in the mid-90's however Li-po batteries, brushless motors, electronic speed controllers and low cost flight control technology were not available at that time.

As aircraft accumulate substantial flight hours, lessons learned such as "mean time between failures" (MTBF) contribute significantly to improving overall reliability. At present single rotor helicopters tend to be a more reliable aerial platform than multirotors, all else being equal, solely because the knowledge and evolution of SR designs has had a several decade head start.

#4 - Small drones fly close to the ground so errors in judging depth perception are common.

When the pilot is inside an aircraft, it is much easier to judge its distance from other objects. For unmanned aircraft flown by line-of-sight (LOS), the drone pilot's eyes are focused primarily on the aircraft in flight and so it is quite easy to lose track of spacing from various objects projecting up from the ground. Also, the accuracy of a person's depth perception drops off dramatically with distance. 

#3 - Small drones aren't built to any formal or recognized quality standards.

As mentioned above, small unmanned aircraft tend to use RC hobby grade components, few of which are built to ANSI, ASTM, ISO or MIL-SPEC quality standards. Sudden unexpected failures can be a frequent event as a result. It is also common to overload systems through trial and error verses having the benefit of accurate operating limits for key components.

#2 - Flying a drone manually is more difficult than flying a manned aircraft manually.

This is primarily due to the fact that the unmanned aircraft pilot is unable to benefit from several key senses, such as sensing motion first hand or detecting changes in vibration for example. When seated inside an aircraft, the relative direction of control remains constant; this is not so for unmanned aircraft. A very common cause for drone crashes is from disorientation following loss of GPS control. When a drone's heading changes, the relative control direction from the pilots perspective can become reversed. The symmetrical design of many multirotors tends to exacerbate this problem.

#1 - Small drone pilots typically aren't formally trained and have widely varying skill levels.

Many aspiring drone pilots think that because the flight controller is doing most of the work that additional flying skills aren't terribly important. This theory only works up to the point that something doesn't go as expected. Being able to react quickly and correctly to an emergency only comes from many hours of flight training, simulation and experience. At the highest level are RC competition pilots who typically have thousands of hours experience building, tuning and flying unmanned aircraft.

There are also less tangible factors that affect the overall reliability of an unmanned aircraft. It is a common mindset that each drone has a number and it isn't "IF" a crash will occur but rather "WHEN". This is a dramatically different mindset from a full size aviator. 

Perfect Perspectives has been able to dramatically improve the reliability of its unmanned systems over the last 10 years by applying concepts such as "all accidents are preventable" along with management of change (MOC) principles and near-miss tracking. Each time a maintenance function, repair or modification is performed, key questions are asked such as - "how might this change unintentionally result in a malfunction?". Going through this thought process has proven very effective in avoiding otherwise unexpected failures. Also any event that is abnormal to the standard flight process is recorded on the Flight Safety Analysis form as a "near-miss". These events are not near mid-air collisions with full size aircraft but rather are any operational close calls that could have caused an accident if not detected beforehand. Studies have shown that for every accident that occurs, there can be 100 or more near-misses that went unacknowledged as a precursor. Taking note of, and incorporating preventive measures for all near misses has proven very effictive in reducing accidents. The combination of all of these techniques has allowed Perfect Perspectives to safely make thousands of flights over several years without incident.