Aircraft Accidents and UAS Data, Part Six

Last week’s web posting addressed the third study Article written by Wallace, Kiernan, Haritos and Robbins, titled: Small Unmanned Aircraft System Operator Compliance with Visual Line of Sight Requirements; the Article was printed in the International Journal of Aviation, Aeronautics and Aerospace, Volume 6, Issue 2 in 2019. This Article, when combined with two previous articles, presented issues with the Unmanned Aircraft System (UAS) and the hurdles to be cleared by the aviation industry to avoid accidents. These facts are long overdue. The aviation community has been placed in a reactive position, which, safety-wise, is not a good place to be.

The unmanned aerial vehicle (UAV) is unique in the autonomous world; they are miniaturized air transportation vehicles. There are fixed wing versions that fly like an airplane and others that operate on rotors. Since UAVs operate within the three axes – X, Y and Z – safety concerns differ from the UAV’s counterparts, e.g. model trains, radio-controlled cars or radio-controlled boats, which pose little threat to the rail, highway and marine industries. To the point, the Rail community is unconcerned with model HO trains breaking free from a basement and causing a 100-car freight train derailment. Yet, a hobbyist’s UAV, hovering near an airport approach, is a real threat to life and safety.

The lesson of First Lieutenant Thomas Selfridge was sobering; aviation was young and imperfect; it was dangerous, yet promising. Once mail was moved by air, we as a civilization could not regress, but improved air mail to become the norm. Pilots still ran out of fuel, crashed into objects at night or were lost in fog. As the DC-8 and the B707 made trans-ocean travel safer and quicker, aviation moved beyond the DC-7 and Boeing 314 Clipper. Older aircraft were pushed out, making way for progress. Still, TWA 800’s lessons reminded us that aviation was still imperfect.

As the UAS industry evolves, we must continue focusing on safety first. My opinion: UAVs are becoming an irreplaceable part of commerce, particularly in the capitalistic environment. Jobs and technological opportunities represented by the UAS are countless; dozens of careers are in the making. However, despite numerous opportunities, dangers are just as abundant. None more critical than the inability for effective oversight.

We are reminded that manufacturers and air operators in the aviation industry have been just as responsible for safety as the Federal Aviation Administration (FAA) has. Decades of regulatory evolution resulted in an oversight system as close to streamlined as can be. The FAA learned to work with the aviation community for the mutual benefit of lowering accident rates and community safety. Is the system perfect? No, but considering the industry’s diversity, aviation is very safe. FAA policies and programs are regularly revised to meet the aviation’s changing demands. FAA aviation safety inspectors (ASIs) are available, seen; they are involved.

Still, the aviation community is concerned, not just with the UAS industry, but with the Congress’s plan to spread the FAA even thinner. To quote JRR Tolkien, “Sort of stretched, like butter spread over too much bread.” To be clear, this is not a small endeavor; the UAS industry is complex, with too many amateurs playing deep below the radar.

Overseeing the entire UAS industry and the aviation community, simultaneously, is impossible. My analogy would be to imagine the Federal Highway Administration suddenly being responsible for the direct policing of every underage speeder, drivers without licenses, fender-bender investigations, reckless drivers, etc., on every road in America while working Monday through Friday, 8:00 to 16:30 local. The FAA is the premier aviation agency in the world. However, is the United States Congress falling on its collective swords putting UAS oversight all on the FAA? Do they understand the magnitude of that responsibility?

To adequately meet the FAA’s needs to oversee both the legacy aviation industry and the growing UAS industry, FAA employee numbers would have to increase four to five times its present manpower. In addition, the FAA’s budget would have to increase considerably, not just to meet the increase in hiring, but to purchase/lease technologies to track the UAVs; renting the available space for attaching the technology to buildings, towers, etc. This data input is crucial to surveillance; it is analyzed to build safety programs for the UAS industry. Another issue: current UAV tracking technologies are manufacturer-specific, e.g. a ‘Brand X’ UAVs can only be tracked by a ‘Brand X’ UAV tracker. Congress would have to budget the FAA with some serious cash to build UAS safety systems, just as the aviation industry uses.

What about the local and privacy laws the UAVs could be violating? How does the FAA square the circle with local law enforcement? Are all laws governing UAVs going to be federal laws or local? Who has jurisdiction? Who prosecutes? Who appears in court? How are UAVs registered? Will all UAVs be required to sport an ‘N’ or equivalent registration number? At present, Title 14 code of federal regulations (CFRs) Part 107 has not addressed this yet.

The Authors’ three Article’s analyzed three unique problems: Article One – pilot awareness of UAVs in their environment; Article Two – tracking UAVs near a small commercial airport; Article Three – potential hazards related to operating UAVs beyond the visual line of sight. The sobering truths of these Articles force the aviation industry to accept undeniable facts: the FAA lacks the technology or the manpower to successfully oversee the UAS industry a-a-a-and the aviation community; large government fails to comprehend the vast scope of this monumental task.

Perspective: A major airline’s FAA certificate management office (CMO) employs roughly eighty to one hundred ASIs, working Monday through Friday, 8:00 to 16:30 in their time zone. These ASIs oversee: thousands of mechanics and pilots; hundreds of aircraft; follow dozens of safety programs, while flying to hundreds of airport ramps – worldwide – 24/7, 365-1/4 days a year. ValuJet Airlines flight 592 crashed on May 11, 1996; ValuJet was strictly a domestic airline. A fundamental reason behind the disaster was the airline’s unchecked growth, e.g. new aircraft and contracting maintenance, that exceeded the ValuJet CMO’s ability to monitor and prevent the safety problems due to ineffective oversight. Twenty-three years later, adequate oversight of the growing airlines continues to be major challenges.

Can the UAS industry receive successful FAA supervision? No. Congress isn’t looking at the reality; elected representatives – with no aviation experience – are relying on the advice of ‘experts’ – with no aviation experience, to make decisions affecting the safety of the travelling public. Congress is pushing the UAS industry off on the only agency that handles aviation ‘stuff’. That’s the only criteria they care about: the FAA handles aviation ‘stuff’.

Can’t the FAA hire more ASIs? First, the FAA hires ASIs who are experienced in their industry to ensure effective aviation oversight. The UAS industry is so young; there aren’t many UAS professionals to hire from. Besides, the UAS industry pays better. Second, the FAA is already seriously understaffed.

The CFRs governing the UAS industry fall under Title 14 CFR Part 107 Small Unmanned Aircraft Systems and its forty-four subparts; these subparts will grow through time and experience. The aviation industry has been regulated under Parts 21, 23, 25, 26, 27, 29, 31, 33 and 35 just addressing airworthiness and certification; Part 43 addresses Maintenance as does Part 65 and 145. Pilots fall under Parts 61, 63, 67 and 68. The certification of air operators’ operations and maintenance programs come under Parts 119, 121, 125, 129 and 135. There are CFRs for schools, medical certification, rotorcraft operations and agriculture. Each Part has dozens of subparts, all based upon different certifications and experiences, e.g. day-to-day issues and accidents.

Can the FAA extend its limited resources to shepherd the UAS industry through its evolutions, and assure the flying public that a ValuJet 592 will not happen again? Will the FAA be guaranteeing safety or effectiveness in its oversight? The US Congress is betting on it. What are they using as ante?

Aircraft Accidents and UAS Data, Part Five

Beginning in November 2016, with Aircraft Accidents and UAS Data, Parts One and Two; then in October 2018 with Aircraft Accidents and UAS Data, Parts Three and Four; I tried to make use of valuable research into the safety aspects of the Unmanned Aircraft System (UAS) vehicles as reported by the Authors: Ryan Wallace, Kristine Kiernan and John Robbins, all of Embry Riddle Aeronautical University and their colleague, Tom Haritos of Kansas State University. Their third Article, titled: Small Unmanned Aircraft System Operator Compliance with Visual Line of Sight Requirements, was printed in the International Journal of Aviation, Aeronautics and Aerospace, Volume 6, Issue 2 in 2019.

To summarize, the first Article dealt with a series of controlled tests that determined if pilots could see, with the naked eye, an unmanned aerial vehicle (UAV) in their flight domain (X, Y and Z axes), even when the test pilots knew the UAV was there. The second article dealt with overseeing different UAVs operating around a commercial airport’s airspace using the latest technology.

The Authors are professionally supportive of UAS success, both in a commercial venue and as recreational devices for, e.g. hobbyists. The Authors are, as demonstrated by their research, dedicated to safety, not only to the UAV operator, but to all the aviation community that UAVs interact with. And that’s the rub: those invested in UAS as commercial entrepreneurs are committed to safety, to expanding their businesses as a benefit, not a hazard. Can the same be said for all UAV operators?

The third Article begins by examining an accident – not incident – involving a small UAS (sUAS) operator and a military helicopter. The September 21, 2017 event involved a DJI Phantom 4 UAV being flown in the evening from as far away as 2.5 miles from the operator, “well beyond visual line of sight (BVLOS)”. The operator flew the UAV solely with a user interface map – no visual contact – into one of a UH-60M military helicopter’s rotor blades. The speed with which a flying helicopter’s rotor blades turn spells disaster when they hit a comparatively stationary object; the damaged blade(s) causes an imbalance in the rotor system, control issues and a loss of lift. The danger exists to the flight crew, anyone flying nearby (‘wing’ man) and anyone on the ground within the crash circle of the falling helicopter. If the out-of-control helicopter happens to be over a city? Numerous fatalities!

The article’s Problem statement: “The potential hazards associated with BVLOS flight represent a clear danger to manned aircraft operators and other National Airspace System (NAS) users. Currently, no data exists to accurately assess the distance at which sUAS operators are flying their aerial vehicles. The authors sought to determine … appropriate waivers or risk mitigation.” A telling Problem statement, especially if one understands the implications. Why is it that “no data exists” in a time when military UAVs could be controlled from across the world for decades?

According to, miniature and micro UAVs have been around since 1990. Remote control aircraft have been around longer, according to, since at least 1985. The technology for controlling a UAV remotely has moved from line of sight to miles from the operator. Why did it take thirty-one years (2016) for industry and the Federal Aviation Administration (FAA) to publish rules for safe UAV operation, e.g. training or operation requirements? Why has it taken thirty-four years for studies to be published that examine the real risks of introducing UAVs into the NAS?

The article’s Purpose statement: “The Purpose of this study was to assess sUAS operator practices with an emphasis on the range and visibility characteristics between the operator and aerial vehicle. This data will be used to establish a baseline of UAS operator flight behavior as well as generate UAS policy and safety recommendations.”

The Purpose statement is also revealing; not just the third Article by itself, but when combined with the first two. These Articles are not anti-UAS; they are anti-complacency. The aviation industry has ignored the UAS industry; UAVs have obviously evolved to the point of major concern. At this time, the aviation industry cannot be proactive, but instead, as always, are reactive. The UAS industry has never been the problem; it was the ‘we’ll-deal-with-it-later’ attitude that has the aviation industry behind the eight ball.

The article utilizes several methodologies to determine operator-to-UAV visual detection efficiency of, e.g. BVLOS and hazards. The Authors state under Hazards of BVLOS Flight that, per a study by Terwilliger in 2012, “Flights beyond visual line of sight have the potential to be particularly hazardous, since they limit the situational awareness of operators. Known as the soda straw effect, the reduced field of view of visual information can diminish hazard recognition and ultimately decrease operational situational awareness.” As I interpret this, the operator controlling BVLOS, stands a good chance of becoming a safety hazard; the operator’s ability to detect hazards is diminished by buildings, natural obstructions, e.g. tree lines or hills, and the distance beyond their visual acuity.

In order to operate a UAV BVLOS, the FAA can issue a waiver to the operator as per Title 14 Code of Federal Regulations (CFR) Part 107.200 – Waiver Policy and Requirements. This waiver, granted by law, can supersede the restrictions of Title 14 CFR Part 107.31 – Visual Line of Sight Aircraft Operation, where the remote pilot, visual observer and/or the controls operator “… must be able to see the unmanned aircraft throughout the entire flight …”

Part 107 is ambiguous at best, about what an operator employing a waiver uses to ‘see’ the UAV’s operating area when BVLOS. Are ‘Brand B’ UAVs only able to be tracked with ‘Brand B’ equipment? If ‘seeing’ is, e.g. a camera mounted to the UAV, can the camera ‘see’ in all directions along the X, Y and Z axes? How would it detect a bird or an ultralight aircraft? How does the UAV avoid hazards or escape becoming a hazard itself?

The Authors’ first article demonstrated the hazard a UAV may present to a General Aviation aircraft under the best visual conditions. The second article established the hurdles of tracking UAVs inserted into a smaller commercial airport’s operating area. This third article highlights the increasing challenges the FAA faces in overseeing the UAS. All three articles increase the urgency of mounting problems, long overdue for addressing. However, one thing the Authors cannot study is the most unpredictable UAS threat to aviation and public safety: Operator Maturity.

For the sake of this article, the subject of terrorist UAV manipulation is off the table; it is a valid argument, but not to the point of FAA oversight and industry coexistence. In the first paragraph of the third Article, the Authors referenced the accident between a UAV and a military helicopter. The operator, per accident report DCA17IA202AB, “was intentionally flying the drone [UAV] out of visual range and did not have the adequate knowledge of regulations and safe operating practices.” This means the operator did not have a waiver to operate BVLOS per Title 14 CFR 107.200.

Is this the first instance of UAVs being operated carelessly through the NAS? No. The FAA has documented hundreds of nightly incident reports where UAVs were operated in the approach and departure paths of commercial airliners at airports, e.g. O’Hare or JFK. These reports date back more than five years. The UAVs, playing chicken with the passenger airliners, were seen by witnesses or flight crews, but it was impossible to locate the UAVs’ operators with existing technologies.

Why would individuals perform these unsafe acts? For years local police in many states have chased individuals aiming laser pointers at helicopters and airliners; these individuals attempted to disorient or temporarily blind pilots during the most critical stage of flight – Landing. A three-year old video showed a ground operator hovering his UAV – BVLOS – directly over the main rotor of a traffic helicopter while it flew over a populated area; the video was captured by a rival station’s helicopter cameras, which then followed the UAV down to the operator, who hid behind trees so as not to be taped.

The third Article is important in understanding why Title 14 CFR Part 107 was written and why these regulations must be applied. UAS individuals are in a larger world, full of situational awareness challenges, where UAV operators are never in any personal danger and won’t suffer from their unsafe actions. This remains new territory, for the UAS industry, the aviation community and the FAA.

Next week we will look at the challenges of oversight by the FAA.

Aircraft Accidents and Lessons Unlearned XXIII: Atlas Air 3591

On February 23, 2019, at 12:45 PM Central time, Atlas Air flight 3591, a Boeing 767-375ER, accident number DCA19MA086, crashed in Trinity Bay outside Houston, Texas. All three persons aboard were fatally injured. That’s it. There is nothing else to report; there is nothing else known.

Yet, even in the early stages of this investigation, there are three lessons unlearned from other accidents that we need to adhere to, to remember. The first lesson, often forgotten, whether intentionally or unintentionally, is compassion; that there are real people hurting because of this tragedy. The second lesson is that speculation is bound to take place; make it count. The third lesson is that, although this was a cargo accident, it deserves the attention of any major or minor accident.

Compassion – One thing I learned early on as a major accident investigator for the National Transportation Safety Board (NTSB) was to look beyond the accident to the families who must now be dragged into the spotlight while their family member’s final decisions are scrutinized by everyone from the pilots’ union to local barbershop regulars. When we allow this type of sensationalism, it destroys innocence, invites doubt and heightens grief. And, it diverts attention from evidence important to finding a root cause. I am reminded of the circuses surrounding such tragedies as Malaysia Airlines MH370 or Germanwings 9535, when the media and their experts felt it necessary to be investigators-without-a-clue and providers of false hope. To everyone within range of my written voice: Let this play out folks, with professional patience, and let the families grieve in peace.

Speculation – there are two types of speculation, good and bad. Good speculation occurs when small groups of professionals, e.g. mechanics or pilots, discuss the accident in the privacy of the line shack or ready room; generating a discussion. There may be those familiar with the B767 or any aviation professional who can glean from what is known about the accident and adjust their Routine – that infamous ally of Complacency. Perhaps a pilot will reread their emergency procedures, again; a mechanic will double-check the maintenance manual paperwork or maybe a pre or post-flight walk-around will take a little longer to complete, allowing the flashlight to closely trace the hydraulic lines in the wheel well. Any boost to our awareness is welcome. Feel free to talk amongst yourselves. Please, keep it professional.

When I investigated China Airlines flight CI611 in 2002, the aircraft still rested on the ocean floor when I arrived in Taiwan. There were no recorders to review yet; they hadn’t been recovered. But, the Taiwanese Aviation Safety Council (ASC) investigators and I didn’t wait for the recorders. Instead, we reviewed all the maintenance accomplished on the accident aircraft, going back as far as we could. We did visual inspections of sister B747s in the fleet and the ASC investigators interviewed mechanics and pilots to see if there was anything to focus on. We were proactive; we speculated in a way that saved time and exploited our resources. That’s good speculation.

Then there is bad speculation. Turning on the evening news to find some self-described aviation expert, proudly holding his model of the accident airplane, maybe with the correct paint scheme, as he waits for the cameras to focus on him. We all know the type: he expounds on opinion with little attention to fact; looking sober while struggling to remember that the model’s vertical stabilizer points up in flight. To the point, this speculation is harmful, not only to the truth, but to the distraction it gives to the investigation. As with Malaysia Airlines MH370, how many experts did it take to make the search crews go round in circles as the recorders’ batteries died? As with TWA 800, how many expert theories did the nation have to listen to, from terrorist bombs to missiles, before the investigation moved forward?

The aviation news website, AvWeb, posted an article on February 26, 2019, referencing a report in Business Insider. Several Atlas pilots – seriously, PILOTS – criticized Atlas for high workloads and low pay. Is this type of selfishness really necessary, to make this tragedy about pay and workload? Perhaps selfies will get these pilots some extra attention. Did their complaining point to root cause or were these pilots just wanting facetime? They presented no evidence; speculation was unnecessary. How do union negotiation issues at this tragic time contribute to the investigation?  They don’t.

Houston’s NBC news affiliate, KPRC-Click2News, posted amateur video on its website: four seconds of Atlas 3591 descending, accompanied by an interview with a former NTSB accident investigator, who was spit-balling without evidence, e.g. “Maintenance issues and mechanical malfunctions, while they are rare events, can and do occur.” What information did that statement provide? None. He then made assumptions about cargo airline work hours, saying, “The effects of fatigue can also be acquainted to someone that is drunk because it impairs not only logical thinking, decision-making, but can have motor skill impairment, as well.” This expert’s logic is all over the place. Maintenance issues? Night shifts? Pilots are, “someone that is drunk”? His dizzying ‘expert opinion’ covered everything but oatmeal food poisoning. Did the ‘expert’ ever fly for a cargo airline? I’m asking because he contributed ab-so-lute-ly nothing to the investigation.

The question of cargo airline flight hours has long been hauled into the accident investigation spotlight; pilot unions, former NTSB investigators and uninformed bureaucrats have lamented the unfairness of operating airplanes in the early morning hours, even though these successful overnight flights, number in the millions since the mid-seventies. I have worked in FedEx’s Memphis Superhub and their Newark Metroplex during those allegedly notorious shift hours. I never observed pilots or fellow mechanics walking drunkenly gate-to-gate; sleeping at the controls or crying out, “Oh, the humanity!” In Omaha, I never saw an exhausted pilot do a half gainer off the crew stairs after his two hour morning flight. As an FAA inspector, I spoke with pilots and mechanics from UPS and other cargo airlines; I never met these lethargic individuals, lurching about beside me as we performed the pre-flight walk-around.

And didn’t the accident occur at 12:45 in the afternoon? Are we ignoring the facts; afraid they will get in the way of speculation and tarnishing the accident pilots’ experience and reputation?

Cargo Accident – I investigated Emery 17, accident number DCA00MA026. Of the Board Members at the time, only Member John Goglia took the investigation the full distance, insisting on a proper hearing. The results of that investigation should have had long term results that should have shaken the aviation community to its core, possibly preventing later air disasters. Why? Because cargo accidents happen in the same airplanes, at the same airports and in the same air space used by passenger airlines; they are affected by the same consequences of that industry.

The NTSB needs to take this accident seriously … very seriously. They need to not only determine probable cause, but they must determine root cause. This accident can’t be about the number of people killed, but about how this affects the entire community; how what is found affects everyone’s safety.

Then there is that four-second amateur video. This video is only a tool, nothing more; a video that supports the investigation, not an investigation that supports the video. The NTSB must not make the same mistake that they made with National Air Cargo flight 102’s amateur video. These videos don’t show the accidents; they show what took place after the accident began.

This accident investigation has just begun. Let us all be professional, compassionate, patient and exercise common sense. We, the aviation community, demand it. The families of the deceased? They deserve it.