Aircraft Accidents and UAS Data, Part Four

The 1960s television comedy, The Addams Family, had several episodes highlighting the family patriarch, Gomez, running his electric trains on intercept courses, accompanied by explosions as the two locomotives ran full speed into each other.  Watching the show, it was something we looked forward to seeing, on the rare occasion it happened.  In those days the term, “unable to look away when witnessing a train wreck” had an expectational gist to it.

These days, that term means something different; to be “It’s like seeing a train wreck; you can’t look away” is a metaphor for: expecting disaster so acutely that, when proven right, it is impossible to avert your eyes – whether physical eyes or mind’s eyes – from the tragedy playing out in front of you.

Last week we reviewed some data recorded in The International Journal of Aviation (IJI), Aeronautics and Aerospace: Evaluating Small UAS Near Midair Collision Risk Using AeroScope and ADS-B.  The paper gave the reader an understanding of how far we have come and yet how far we have to go regarding small unmanned aerial system (sUAS) platforms – hobby or toy aircraft.  One point the paper made in its first pages was pivotal to answering a question: Who will assure safety with sUAS platforms?

The paper lists an event where a US Army Sikorsky UH-60M was hit by a Phantom 4 sUAS (National Transportation Safety Board (NTSB) Accident number DCA17IA202) while the helicopter was part of military operations being conducted near Hoffman Island, NY.  The helicopter suffered damage to its main rotor, but landed successfully; the Phantom was destroyed.  The NTSB identified the sUAS by its unique serial number found in the wreckage; the FAA then traced the sUAS back to the owner.  The probable cause of the accident was “the failure of the sUAS ‘pilot’ to see and avoid the helicopter due to his [the sUAS operator’s] intentional flight beyond visual line of sight.”

The paper stated in a section titled ‘Problem’, that, “That the Federal Aviation Administration (FAA) has undertaken varied efforts to contain the problem of unsafe or non-compliant sUAS operations.  From June 2007 to May 2018, the FAA pursued [enforcement] action against 518 sUAS operators;” this information was provided by the Government Accountability Office (GAO).  The IJI article further stated that, “In a rebuke to the FAA, the GAO concluded that FAA safety efforts are hindered by a lack of reliable sUAS operations data.”

The GAO is part of the Legislative Branch of the United States government; it has responsibility for auditing the other Departments that answer to Congress.  Although providing a means to assure the different Departments are meeting their assignments and are properly funded, they do not specialize in areas, e.g. unmanned aerial vehicle (UAV) platforms.  To this point, the GAO states in the previous paragraph, that the FAA has been pursuing actions against sUAS operations since 2007 to present.  For the first nine years of that time, the FAA has restricted its UAV attentions strictly to the safety aspect.  The FAA had not been pursuing oversight of UAVs since there were no regulations or policies that gave them authorization.  Instead the FAA was trying to prevent UAVs from having accidents with manned aircraft, e.g. following up on reported near misses, UAV strikes and UAV sightings where trespassing in the National Airspace System (NAS) has occurred.

Yes, there is a difference.  The regulations for UAVs relating to the unmanned aerial system (UAS) are only two years old; their introduction was June 28, 2016.  The time needed to write Title 14 of the Code of Federal Regulations (CFR) Part 107 – Small Unmanned Aircraft Systems, took between three and five years to author the forty-four subparts – 107.1 through 107.205 – that are listed today.  For the time the FAA and industry wrote the regulations, the FAA was not regulating the UAS, while they were regulating the airspace they were being operated in.

I stated last week that the IJI paper is valuable beyond the data collected; it points out a serious flaw in the direction UAS oversight and safety is going.  Ironically in 1958, if the Department of Transportation could have foreseen the rise of the UAS, would they have trusted the FAA with oversight of the program?  My guess would be they would not.  It is not due to competence, but to several aspects: priorities with manned aircraft, technology, finances and manning.

To be clear, the FAA is chiefly responsible for manned flight; that is their priority: aircraft from crop dusters to airliners, from tour helicopters to oil rig helicopters.  The oversight of unmanned vehicles is, frankly, too far beyond the FAA’s capacity as it stands today.  Unfortunately, being the regulator of the skies, the FAA has been volunteered without a say; painted in a corner; victims of their own success.  Metaphors aside, the FAA is out of their depth.

The technology is not there yet, even for collecting data on near misses or trespassing inside airport perimeters.  The push for the FAA to register sUAS platforms was put off for months as lawsuits successfully stymied the program; the lawsuits were later reversed, the registration program was found to be justified and started up again, but at a loss of time.  In the IJI report, the research team employed a Dà Jiāng Innovations (DJI) Aeroscope, a high tech sensor that finds sUAS … but only DJI sUAS.  Does the technology exist to find all brands of sUAS?  Will the technology be able to not just identify any sUAS platform, but determine its speed, direction, altitude, size, threat, owner, etc. effectively?

Speaking from a finances perspective, how will this technology be paid for?  Will DJI Aeroscope devices need to be leased or purchased?  How many would be required per square mile?  Will building owners allow an Aeroscope to be mounted or will they charge rent?  Who pays for these things?  And how many inspectors would need to be hired to meet the sUAS oversight need?

That last question brings us to manning.  Newly-hired aviation safety inspector (ASI) training classes have dropped dramatically in number over the last few years; hiring has been mostly dedicated to replacing the retiring ASI workforce, to stay in front of growing air operators and an increasing number of certificate holders.  Certificate management offices and flight standards district offices are dedicated to manned flight; they are already outnumbered with regards to manned oversight and surveillance.  Finding qualified UAV-experienced ASIs who can adequately perform oversight is yet to be put into effect; the future sUAS ASIs are simply not out there.

Previously, changes in the industry have been incremental.  From propeller-driven intercontinental aircraft to jets; from four-pilot to two-pilot aircraft; from analog technology to digital technology, the FAA’s many divisions have been able to stay in front of the times.  The UAS, however, is vastly different.  Mechanically speaking, there is no basis for how long a UAV can operate before heavy maintenance; what constitutes maintenance or even what systems will be important – don’t forget, no pilots.  From the Operations side, are they certificated like a pilot if an sUAS is operated in the NAS?  How will they be drug tested, according to what drug program and how is that enforced?  Does Little Johnny have the same restrictions that Captain John has when operating an sUAS around manned aircraft?

Most importantly, how does the FAA assure every sUAS is operated to the highest level of safety per Title 14 CFR Parts 107, 91, 43, 65 and so on?  How does the FAA get its arms around the rising number of UAVs and their operators?  This is the conversation that the aviation industry should be having.

 

This wave could not have come at a worse time in history.  America’s complacent belief that we have the safest aviation system ever is a myth; it is a false illusion.  Safety is not guaranteed by achieving a goal but by constantly raising the bar.  However, organizations, e.g. the NTSB, will run with the ‘Skies-are-Safe’ routine because it gives the flying public a warm feeling; it’s not necessarily true.  The ability to regulate the airspace of the size and complexity of the United States cannot be compared to the safety needs of an airspace over, e.g. France, Great Britain or Japan.  There is no comparison, yet aviation safety professionals make the comparison, nonetheless.

If the industry chooses to ignore the UAS wave overtaking them, then they, too, fall into complacent denial; they are renouncing the pending perfect storm.  Complacency is like a train wreck; one cannot look away.  Instead one becomes fixated on what should have been done differently when there was time.  All the while, the tragedy unfolds in slow motion.  The wakeup call is coming; it will be devastating.

Aircraft Accidents and UAS Data, Part Three

On November 17th and 24th, 2016, I wrote two articles titled: Aircraft Accidents and UAS Data, Parts One and Two, respectively.  My article’s information referenced research conducted into the safety of Unmanned Aerial System (UAS) vehicles.  The much anticipated follow up journal paper has recently been published by the team of aviation authors: Ryan Wallace, Kristy Kiernan, Tom Haritos, John Robbins and Godfrey D’sousa, all hailing from the respectable Embry-Riddle Aeronautical University (ERAU).  They have published the second journal paper in the International Journal of Aviation (IJI), Aeronautics and Aerospace, titled: Evaluating Small UAS Near Midair Collision Risk Using AeroScope and ADS-B.

https://commons.erau.edu/ijaaa/vol5/iss4/2/

The team of aviation authors’ first IJI article from 2016 placed a Cessna 172S G1000 aircraft in ‘harm’s way’ in a controlled environment by having a test group of different aged pilots identify an unmanned aerial vehicle (UAV) while conducting normal flight maneuvers.  One aspect of the tests was that the pilots expected the UAV in their midair neighborhood, but were not told of where.  The results were alarming, not in pushing any narrative that suggests UAVs and public aviation don’t mix, but in that the UAVs were still difficult to perceive (detected between 0.10 and 0.31 statute miles away), identify and react to their presence, e.g. avoiding close encounters.

This second IJI article was more aggressive and, in this writer’s opinion, more telling, in the information they presented.  What must be distinctly understood is that this group of researchers are pro-unmanned aerial vehicle technology and use – as am I; my interest being from a business standpoint.  Indeed, it is the authors’ commitment to UAS industry success that gives them a unique position that can promote the advantages while simultaneously strongly warning of how UAV platforms, particularly small unmanned aerial system (sUAS) platforms, can be misused by those without the proper discipline.  As this writer interprets the term: ‘sUAS’ is (in the IJI report) an amateur-operated UAV; it is not operated professionally by an approved business interest, e.g. real estate firms, marketing firms, professional photographers or agriculture consultants.  However, the sUAS platform is large enough to hinder the safety of the National Airspace System (NAS) and small enough to be practically undetectable with the human eye at aircraft-operated speeds and technological identifying means.

There were four research questions the team sought to answer:

  1. “What are common characteristics of sUAS flight locations?”
  2. “What are common characteristics of sUAS operations?”
  3. “What is the potential impact of detected unmanned activity to aerodromes and aviation operations?”
  4. “How effective are geo-fencing restrictions in preventing sUAS flights from entering protected areas?”

The author team employed an applied research method; they utilized “exploratory research and case study approaches.”  To do this the team had access to a Dà Jiāng Innovations (DJI) Aeroscope mounted to the top of a three-story education building near Daytona Beach International Airport (DAB); the DJI Aeroscope, as per DJI’s website: www.dji.com/aeroscope “is able to identify the vast majority of popular drones [UAVs] on the market today by monitoring and analyzing their electrical signals to gain critical information, allowing users to protect the integrity of their flight-sensitive environment”.  The Aeroscope was used to detect small unmanned aircraft near the airport for a thirteen-day period.

Even with this advanced means of detection, this equipment was fairly limited – technologically.  The DJI Aeroscope only detects DJI-manufactured sUAS platforms; the Aeroscope detects only platforms within electronic line-of-sight; some DJI platforms could not be identified and the authors did not assess environmental or seasonal factors, e.g. weather conditions.  However, DJI has a large market share of UAV and sUAS sales (72%), which increases the finding’s credibility.

When detecting these platforms, one must remember that this is a three-dimensional tracking.  For example, when observing proximity to an airport runway, the X-axis is the straight line distance to the end of the runway; the Y-axis would be the straight line distance to either side of the runway.  Just as critical is the Z-axis, representing the vertical space above the runway that intercepts an aircraft’s flight path on take-off, approach or turning.  By contrast, the sUAS’s owner has the limits of a human eye’s view of, e.g. a runway; the sUAS can see straight down the runway, but fail to see the aircraft approaching from behind.

The report opens with a February 2018 event: an sUAS’s video footage hovering behind a Frontier Airlines A320 on approach to McCarron airport.  This sUAS, hovering in the space that the A320 passed through is disturbing enough, but it is also unable to ‘see’ any other aircraft following the A320’s flight path into the airport.  How dangerous is an impact between an sUAS and a major aircraft?

The University of Dayton (UoD) released a video of a controlled collision, showing a 2.1-pound sUAS impacting a Mooney M20 wing at 238 miles per hour.  Please review the video:

https://udayton.edu/blogs/udri/18-09-13-risk-in-the-sky.php

What must be taken away from this experiment is that the M20’s curved leading edge is structurally stronger than a flat panel would be; despite this fact, the sUAS disappears into the leading edge.  Behind the leading edge is one of the aircraft’s fuel tanks.  Is it unlikely that one sUAS can hit the narrow profile of an aircraft’s leading edge?  Yes, I repaired a B727 that ran into a flock of geese that structurally damaged six separate points on both of the B727’s wings’ narrow leading edges; four slats and one Krueger flap were replaced because of the extensive damage.

In addition, even with the damage incurred by the UoD wing, the aircraft might – might … still fly to an emergency landing; remember, an sUAS is electrically powered entering a fuel cell, while a goose is not.  However, the same sUAS hitting the turning main rotor of a helicopter would be catastrophic.  The helicopter is solely dependent on the main rotor system to remain airborne; each main rotor blade must be balanced with its opposing rotor blade(s) to maintain flight; if even one blade is damaged or shattered by a similar sUAS impact, the helicopter will not stay in the air.

The authors found sUAS activity came as close as within a half mile (X- and Y-axis) of DAB’s location; some sUAS were found to be flown above 1000 feet (Z-axis).  The report shows high-altitude sUAS flight activity along nearby Ormond Beach, where patrol helicopters and banner towing operations are frequently found.  The report states, “These findings demonstrate that at least some sUAS operations are penetrating altitudes traditionally reserved for manned aircraft operations.”

The effectiveness of Geo-fencing was reviewed.  Per the report, “Geo-fencing is one or more location-specific, programmed flight restrictions or limitations designed to prevent or restrict sUAS flights over or near areas that would create a security or safety risk.”  There are four categories of Geo-fencing zones: Warning, Enhanced, Authorization and Restricted; each is designed to alert sUAS operators of flight restrictions, yet each restriction can be overridden depending on the operator’s unlock capabilities.  It is unclear if aircraft operators receive a similar warning of an sUAS trespassing in the area.

In defense of sUAS operations, the report states, “This data suggests that a preponderance of sUAS operators are flying for personal use around their own residences.”  The authors then comment, “The authors were particularly concerned that nearly 97% of all detected sUAS flights had been conducted within five statute miles of one or more aerodromes”; an aerodrome is defined as a small airport or airfield.

The IJI article’s authors are also taking a hard look at a smaller international airport: DAB, whose number of operated flights are considerably less than Kennedy, Los Angeles or O’Hare airports.  These three major airports – and many more – have seen an increasing number of sUAS sightings in close proximity to the approach paths and altitudes used by major airliners.  DAB, for its size, is demonstrating an increasing congestion of sUAS activity, as is its surrounding area.

The IJI article is a snapshot.  It does not definitively say one group, sUAS operators, are a danger.  However, the sample population data presented suggests that the general public has a long way to go in cultivating a safe environment for the flying public, that the data cannot be ignored or trivialized.  It truly is no longer a matter of ‘if’, but a matter of ‘when’.

Next week I will revisit this IJI article to discuss concerns the report raises about policing our NAS.  I highly recommend that readers review the IJI authors’ report.

Aircraft Accidents and Lessons Unlearned XVIII: AirNow 352

There is an Aesop Fable titled, The Boy Who Cried Wolf; it is a story about a shepherd boy who twice tricks the local villagers to come running to help him by falsely crying an alarm of “Wolf!  Wolf!”, thus amusing him.  When the real wolf did show up, his cries for help were ignored, leaving the villagers to say, “We will not be fooled again.  Pah!  If there is a wolf, then let the animal satisfy itself on mutton and boy.”

The last statement (“… then let the animal satisfy itself on mutton and boy.”) acts as a metaphor.  The boy – the National Transportation Safety Board (NTSB) – has been crying “Wolf!” for years; they do this when they cry, “Pilot Error” or “It’s the FAA’s fault”.  This is indicative of indolent investigating; ‘phoning in’ the Probable Cause to avoid investing more time or energy.  What results?  The Industry and manufacturing do not take the NTSB seriously – or serious enough – to warrant further effort.

The AirNow 352 accident investigation demonstrated the NTSB’s inability to understand the Industry, Manufacturers and the Federal Aviation Administration (FAA).  However, this was not just a failing on the NTSB’s part; unlike the NTSB, the FAA does not have the luxury of Industry ignorance.  Their weak response is inexcusable.  This accident is indicative of how politics destroys the FAA’s effectiveness.

Business Air, Incorporated (as AirNow) flight 352 was an Embraer 110P1 flying under the provisions of Title 14 Code of Federal Regulations (CFR) Part 135: Air Taxi and Commuter, from Manchester, NH (MHT) to Portland, ME (PWM) on November 8, 2005.  The cargo aircraft was taking off out of MHT when the number one (left) engine failed.  The pilot could not stop the aircraft from making a shallow left turn; the aircraft crashed into a Walmart garden center.

The Probable Cause of the accident final report for Accident Number NYC06FA027 read:

“The pilot’s misapplication of flight controls following an engine failure.  Contributing to the accident was the failure of the sun gear, which resulted in the loss of engine power.  Contributing to the sun gear failure were the engine manufacturer’s grandfathering of previously recommended, but less reliable, maintenance standards; the FAA’s acceptance of the manufacturer’s grandfathering; the operator’s inadequate maintenance practices and the FAA’s inadequate oversight of the operator.”

The pilot is accused of misapplying flight controls following an engine failure; it would take a pilot to verify that allegation.  The allegation is another way of crying “Wolf!” i.e. ‘Pilot Error’.  However, as will become evident, the questionable FAA oversight would suggest: inadequate training of the pilot and/or operation of the equipment (Embraer 110P1).

The Probable Cause then spoke to maintenance.  The first-stage sun gear did fail, mainly because it was reused beyond its physical lifetime, instead of being discarded after it exceeded its life cycle, it was removed and then … reversed – placing it back in position, to wear the opposite side of the sun gear’s teeth.

An impact on the sun gear failure was the engine manufacturer’s – Pratt and Whitney Canada (P&WC) – grandfathering of previously advocated maintenance standards, i.e. Business Air’s On-Condition program.  NOTE: The term ‘grandfathering’, as per the Merriam and Webster Dictionary, refers to ‘permitting to continue under a grandfather clause’.  In the case of AirNow 352, overhaul times were escalated based on Business Air’s unreliable maintenance standards per dubious data from the On-Condition maintenance program.

An On-Condition program means, “Engine manufacturers provide a time limit for when their engines should be overhauled.  This limit, which consists of a certain amount of Flight Hours and Calendar Months, is the point at which they recommend the engine should be removed from the aircraft and sent to the factory/engine shop for overhaul.”  Because the manufacturer, P&WC, grandfathered the overhaul requirements of AirNow 352’s number one engine, the sun gear was not replaced for 22,000 hours instead of the recommended 12,000 hours.  Furthermore, it was not replaced, but reversed.

However, it is the last independent clause (italicized) in the Probable Cause that is the cry of “Wolf!”  Like the term: ‘Pilot Error’, the NTSB used the ambiguous term: ‘the FAA’s inadequate oversight’, too often.  The phrase becomes an overused string of buzz words without supporting evidence.  What about the FAA’s oversight was inadequate?  Where did they fail?  Because of these “Wolf!” cries, there is no urgency in finding solutions to real problems.

The AirNow 352 accident was indicative of what could and would go wrong with the FAA’s oversight of an oft overlooked operator.  Despite Business Air falling under the radar with regards to popular air operators known to the public, Business Air, AirNow’s parent company, was located in Bennington, VT, near the western border of the state.  The Certificate Holding District Office (CHDO) responsible for Business Air’s certificate oversight was located in PWM, a full workday’s drive from Bennington.  Each time the Principal Inspectors conducted on-site surveillance of Business Air, the travel expenses included four nights for hotels and per diem.

In contrast, the Flight Standard District Office (FSDO) in Albany was located less than an hour from Bennington, VT, and Business Air; its location would have provided better oversight and more frequent on-site inspections.  However, the politics of office funding for CHDOs and FSDOs is dependent on their assigned certificates and territory; the quantity of certificates versus quality of certificate management.

The Principal Maintenance Inspector (PMI) assigned to Business Air at the time of the accident was temporary, swapping out with other PMIs on a rotational basis, changing one Principal for another.  This guaranteed no FAA Principal consistency, resulting in bad decisions; unfamiliarity due to a complete lack of understanding of Business Air’s culture.

As mentioned previously, the engine overhaul times were being grandfathered by P&WC based on unreliable data from Business Air’s engine monitoring program; as Business Air applied for extensions to overhaul times, the data they provided was not accurate.  Each FAA PMI, having just come on to his/her assignment, did not get a consistent picture of what Business Air was doing, blindly giving maintenance exemptions based on unfamiliar data.

It is important to note that Business Air was a Title 14 CFR 135 operator; the aircraft was single pilot hauling cargo.  There are no headlines here; the pilot survived, they hauled cargo and no more than one pilot flew at a time.  In the world of public interest, this was no more than a shrug of the shoulders in the news.  No one outside of New England, indeed, Bennington, Vermont, would have been aware of the accident past the second week and even less interest as time continued to pass on.

However, this was not the only accident Business Air’s pilots had experienced; there were three others, two of them fatal.  One Business Air aircraft flipped on the runway during approach to an airport; a second one slammed into the side of a mountain near Bennington while approaching at night; the third was forced to land in a field because of fuel starvation when the pilot did not request enough fuel.

Four accidents demonstrated that Business Air’s oversight was lacking.  Fueling and flight training all fall under the purview of the Principal Operations Inspector; engine overhauls are the responsibility of the PMI.  Substantial breaks in the surveillance practices and continuity of oversight of Business Air could arguably be the cause of miscommunications between the Operator and the managing office.  It also demonstrates why operators cannot be trusted to self-police their certificates, especially when they are considerably small, out of the geographical driving distance of the CHDO and can exploit the unfamiliarity of the Principal Inspectors with the operator’s culture.

AirNow 352 is the perfect example of opportunity ignored, both by the NTSB, who spent too much time making accusations to recognize the real deal when it presented itself; and the FAA, who discounted their weaknesses (consistent oversight of Business Air) to maintain a feeble grip on Business Air.  Since November 2005, nothing has been done to rectify the omission.  It’s what happens when you cry, “Wolf!” too often.