FAA, flight checks and drones

I spent most of my FAA career in field installation work.  The last major project I worked was a new Airport Traffic Control Tower at Broomfield, Colorado.  This was to replace an existing ATCT that was poorly positioned and too short to view one of the entire runways.  I had only a portion of the engineering work assigned to me, which was installation of communications and RADAR.  The RADAR work there consisted of a display system that is fed out of Denver International Airport.  The project was a great success for all involved.

During the review of the antenna placement on the roof, I was concerned about the possibility of multipath because the air-to-ground communications uses a very old form of modulation, AM, that is particularly susceptible to multipath.  Multipath is a condition where the signal transmitted by an antenna is reflected so that there are multiple transmission paths of differing lengths between the receiver and transmitter.  So, the receiver senses interfering unsynchronized signals and the result can distort, weaken or entirely nullify communications until the aircraft moves into a place where a multipath problem does not exist.

Multipath is not an usual problem.  It is hard to predict exact locations where it might occur.  Controllers typically discover where the problems are and if it happens in a really bad spot (like final approach), technicians are asked to fix it.  I thought that it would be a much better idea to probe the airspace and find the multipath zones before the ATCT was commissioned.  For this, an extensive flight check would be required.  I contacted the Flight Inspection Field Office (FIFO) and spoke with several people about this.  I had used their services many times in the final checks of new RADAR facilities, but never for probing multipath zones.  One FIFO pilot was particularly helpful and told me he had made a check like that.  I asked him how it was done.  It was like the cell phone commercial, "Can you hear me now?"  I was really not surprised.  

Of course, cell phones companies do not probe their cellular coverage as those commercials depict.  They use automated equipment like the Anritsu 2713E.  This unit has the ability to produce a KML file, which is extracted to a thumb drive.  KML script opens Google Earth and directs it to plot the data with various features.  In this case, the data is a time- and position-stamped point that appears as a colored dot on Google Earth.  Click the dot and a window opens with exact measurements at that point, along with information about the instrument condition when the data was recorded.  The readings are separated by about one second.

I contacted Anritsu to discuss the possibility that this cellular network coverage mapping instrument might be adapted to perform flight checks.  Eventually, this aroused enough interest at Anritsu that they sent a team of four engineers to my office to discuss it.  Within FAA, I could arouse almost no interest.  I was the only government employee who attended the meeting.  It was not hard to convince the Anritsu engineers that this was an excellent idea.  They were excited because the potential market for such a device would be considerable and give them excellent material for advertising, I suppose.  They agreed that it was likely that FAA management would be willing to adopt such technology, once it was adequately demonstrated, so they made the decision to build software to meet my specification. It had to work with AM radio and it had to produce 3D data.  The development of a flight check instrument was done at no cost to the government.

After a few months, they had made the software modifications and it was installed in a 2713E unit that I had purchased using project funds (with the permission of the Project Engineer).  I tried it out as described in this final official report.  It worked beautifully.  With a few seconds of viewing the Google Earth projected data, one could confidently ascertain the weak locations of communications coverage for a given frequency and ground antenna location.  This could be done on the fly, so to speak.  The data could be evaluated during the flight check and the ground antenna adjusted in order to repeat the test without even landing.

Making a flight check like this had big advantages.  It required the use of only a small aircraft, locally procured.  Many FAA employees are pilots and small aircraft can be leased at low cost.  The opportunity to log some hours for free, and even get paid doing it would generate volunteers easily.  The modification to the aircraft was coordinated with Flight Standards and it was trivial, basically unscrew the port to one of the external antennas on the aircraft and attach the Anritsu.  As long as the plane had two communications antennas, it would work.  I rented aircraft on two occasions and made extensive recordings as seen in the official report.  The cost was trivial compared to what FIFO would cost, as they use large twin engine aircraft with a crew of up to four that flies out from Oklahoma City and stays overnight usually.  I might estimate the cost savings to be 95% and the results were vastly better than any flight check using the "Can you hear me now?" method could accomplish.

The public popularity of drones was growing and the technology was getting better and better.  FAA was directed by Congress to accommodate the growing demand for controlling them.  There was a special office set up in DC for this purpose.  I contacted the director of this office to make a suggestion.  The modifications and addition of equipment and procedures for handling drone (Unmanned Aerial Vehicles, UAV) traffic was his responsibility.  He was really focused on developing new procedures.  He sent me an organization chart that was mind boggling in complexity.

I suggested that the use of a UAV for official flight checking was in the interest of the government.  FIFO flight checks are very expensive and flight crews can be hurt or killed in the event of a crash.  Flight checks routinely involve routes that are quite close to the ground, including over populated areas, such as the one I had flown to test the Anritsu (which probably generated a few calls to the authorities).

One might get the impression that a twin-engine aircraft is safer than a single-engine because it has redundancy, and this is not wrong, but it is not really a simple matter.  If an engine stops operating, it changes from being a source of thrust to being a source of drag.  The effect upon the aerodynamics create an emergency that must be correctly met by the pilot very quickly or risk fatal consequences.  If a single-engine loses power, it becomes a clumsy glider and control of the aircraft is not difficult, although finding a good spot to land or crash is then the priority.  In either case, if the engine loss occurs at high altitude, the options are much better than if flying close to ground, such as in most flight checking.  Piloting a twin-engine near the ground and losing one engine is almost certainly to result in a crash with no control, so the possibility of large loss of life is significant.

The UAV flight check aircraft would weigh a small fraction of a FIFO aircraft.  It would be flying as slow as the operator desired, so as to gain a very fine data resolution.  It could have many thrust sources, so a loss of one would be uneventful because of automatic compensation.  Good places for emergency landings could be previously mapped for automatic crashing or landing to avoid casualties.  Even if it did crash into a populated area, the weight of the UAV and speed would mean the damages would likely be relatively trivial.  Air-to-air beacon transponder interrogation (TCAS) and other collision avoidance techniques would be required to minimize danger to other aircraft.  A UAV could be designed so that it was virtually impossible to intentionally hit it with another aircraft because they can be so maneuverable and the sensing technology is advanced.  This is a lot easier of a challenge than a self-piloting car.

My suggestion pointed out the advantages of adapting the air traffic control procedures and equipment using government controlled UAV.  This adaptation must be done safely.  There is no way that a new set of procedures can be produced by some people sitting at desks that can just be introduced one day that safely allows the Oklahoma Land Rush of increased drone activity.  It has been many years since the FAA was mandated to adapt the National AirSpace (NAS) to allow drone flights safely.  It has not happened, although we frequently are told that it is coming soon.  It may be that this delay is due to this initial problem of sudden transition.

What is typical in government bureaucracies is that a difficult challenge is met with delay and obfuscation, accusations and recriminations.  The goal then becomes to retire before having to actually deal with the challenge.  The use of outside contractors like Mitre, Raytheon, or Lockheed acting as consultants to apply specialized knowledge is common.  The contractor then hires many retired FAA managers and controllers to apply their expertise in developing a solution.  The chief advantage of this, as the FAA Inspector General noted, is that it moves responsibility outside the agency.  If the boondoggle fails, it is written off as a need to find a better contractor and then repeat the process. 

This is how the bureaucracy works.  They see every challenge as an opportunity to build their kingdoms.  Two engineers I know were approached by air traffic management during the planning of the new Denver airport.  The AT managers wanted them to build a system that would allow rapid reconfiguration of air traffic patterns.  There are a number of standard traffic patterns that change with weather and traffic demand.  When a change of configuration is made, it can really put the controllers under pressure to rapidly reconfigure the field equipment, and there is no room for error.  AT wanted to be able to have a computer screen where they could make one click and change to a different configuration automatically.  The engineers could see this would be a challenge, and they also knew that if the DC office found out that they were doing this, it would be killed.

There had been a major program before I was hired in 1986 called Advanced Automation that included a reconfiguration system like this, but it utterly failed.  The Inspector General's report in summary stated that they wasted $6 billion with nothing to show for it, except for a few obsolete office computers.  The IG report documented communications among FAA management in which it was obvious that no one had a clue what to do, and they did not care that they did not know.  They hired IBM and IBM hired a lot of FAA retired managers at high pay to brainstorm solutions.  After a few years of doing very little, the whole program was shut down.  It was a huge embarrassment and Congress threatened to privatize the FAA.

So, AT was frustrated and knew that the local engineers could take care of developing a reconfiguration system for their new airport, which they did.  Denver became the first airport to have this reconfiguration system called ICMS (Integrated Control and Monitor System).  Other AT managers at other major airports wanted it, so these engineers build them, eventually for 17 major airports, at a cost that was in the low million$.  I know this because my cube at Denver was adjacent to one of the engineers and I was well informed.  Headquarters was very unhappy about this loss of a potential boon for their bureaucratic bloat.  They refused to recognize the ICMS or allow it to become a certifiable system.  In my view, this is an abrogation of duty.  Certification acts as a threshold of minimal quality of operation.  Technicians must have a set of standards below which the system must be removed from service and repaired.  Without being recognized as a certified system, ICMS remained in a limbo where spares, training and procedures were not handled like normal.

The bureaucrats see the need for a new system as opportunity for massive contracts that are obscenely inefficient and deliver inferior products.  A defective product creates a gravy train that can run for the entire life of the equipment.  It is showcase corruption.  Creating a false rush to meet schedule is very important because it creates wide implementation of systems before they are fully designed, then years of going back to re-engineer every site.  Change Orders are extremely lucrative.  We must expect that a new system will require changes, as these are quite complicated systems and the total number produced is small, but the procurement system has obviously been gamed for massive illicit payoffs.  There is plenty of precedent for this in government contracts of very many kinds.

The best way to adapt the NAS to UAV traffic is with a small number of UAV on a trial basis and a dedicated site to adapt the NAS equipment and procedures by a few engineers and technicians who specialize in the required systems.  The big one is the RADAR Automation System.  UAV must have beacon transponders, which provide precisely accurate position data, velocity and identity information to the controllers.  The RADAR Automation would be modified to allow for direct control of each UAV by an FAA controller.  A grounding function that would allow a controller to clear an area of concern of any UAV traffic if, for example, might be implemented.  If traffic was building on some runway approach end and the controller decided that UAV traffic was an unnecessary distraction, this could be done automatically and without a need for cooperation from the UAV pilot.

Identify Friend or Foe (IFF) is Mode-4 of the beacon interrogation system, going back to very early beacon RADAR development.  If an aircraft comes into a protected airspace, a Mode-4 interrogation is triggered.  If the aircraft does not reply with the code of the day, the aircraft is a candidate for being an enemy.  The same kind of thing may be necessary for drone traffic.  We may be facing a future where a lot of sensitive facilities are equipped with anti-aircraft weaponry that can handle a large swarm of UAV that transgress secured airspace.  UAV are small and usually depend on either GPS or remote control for delivering a bomb payload.  Firing intense microwave bursts or laser may be sufficient for economical defense with minimal collateral damage.  Anti-aircraft missiles cannot be wasted on drones because of attrition.  An enemy could simply keep flying cheap drones into the airspace until the defenders run out of missiles, which would not be very long.

The development of a whole new branch of air traffic, the public UAV sector, needs to be approached by such a gradual method to minimize risk, while achieving the best service to the public UAV community.  By starting with a small number of government controlled UAV, these policies and equipment changes could be implemented gradually, first at low traffic density airports, and then onto more difficult airports.  When the required system changes are discovered and implemented, then begin to allow public UAV traffic at the low density airports and gradually increase implementation across the NAS.  Of course, this has the massive advantage of modernizing the flight check method and system, a move that would save enough money to likely pay for the whole FAA UAV effort, and then some. This was my suggestion to the DC UAV office.  It was reflexively rejected without any review.

The government UAVs used for this flight control procedure and equipment development would be used for flight checking facilities, which would create a large savings in maintenance costs, and a big improvement in maintenance quality, while reducing risk to the public from a possible FIFO plane crash.  The UAV office manager was not interested.  He repeatedly told me that "the decision had been made".  He refused to discuss the great advantages offered by the approach I was suggesting.  This was before anything other than a few flowcharts had been produced.

I reported all this back to my local chain of command.  My immediate supervisor gave me an award.  A senior engineer told me to "cease and desist" from any further activity on this project.  I asked why.  He told me that it was "illegal".  I asked what order I was violating.  He said that he did not know, but he was sure it was a violation.  He referred me to another engineer, who I contacted.  That engineer had apparently tried to develop something similar many years earlier.  That engineer's replies were evasive.  He just hinted that it would not end well if I kept pushing it.  Another engineer told me I would be fired.  A friend who was raised in a communist country was fearing for my life.

I contacted FIFO, NTSB, Flight Standards, FAA Technical Center and congressional representatives explaining my situation and tried to find any regulation that might prohibit such development or use of new instrumentation.  They just referred me back to my FAA chain of command, and they got some laughs.

Experiences like these convinced me that government does not exist for the welfare of the population.  This is not entirely true.  The controllers and technicians are the ones really dedicated to keeping the flying public safe.

Can I believe that government is actually going to investigate new discoveries in physics and make sure that possible advances that would dramatically improve life on Earth will be sought out, even if such changes would reduce the power of the powerful?  No, I cannot believe that government is capable of operating in the interests of humanity, generally.  Government operates in the interests of government, which may explain why governments grow corrupt and collapse.  As the saying about the mafia goes, "The fish rots from the head."  Political solutions are the clean up operation after the mob has taken over another section of human activity.

Remember, the prudent threshold of evidence required to justify funding an effort to determine if hydrinos exist or if the excess heat has any commercial utility is only, 'is there any chance that it is real'?  This is because if it is real, it would be of such tremendous benefit to all of life on Earth, and the means and method of answering that question are already readily available in many labs, inexpensive and quick. 

The Anritsu is fairly small and lightweight and can be controlled remotely.  The military is obviously producing UAV pilots with great expertise, as well as surplus heavy payload UAV.  The FAA is supposed to preferentially hire veterans.

There is no question that the conventional method of performing flight checks is very expensive.  A thorough flight check of a long-range RADAR, back when things were cheap, was over $200,000.  A flight check has to be reserved long in advance, so weather or delays caused by problems elsewhere can create a postponed flight check.  Mechanical problems were very common, grounding planes with no warning.  This means that scheduling the flight check was hit-or-miss and at remote facilities (there are thousands of them), a crew of technicians had to remain ready to assist at a moment's notice possibly for days, and this gets very expensive.  

What does this have to do with cold fusion, Randell Mills, hydrinos or science?  It is a clear example of a failure of government to adapt to a challenge created by a minor advancement in technology (UAV flight checks) which created incoherence, economic waste, and unnecessary risk.  In my long experience with the FAA, which is reputed to be one of the most efficient and well run agencies, this kind of failure is very common.  Is it reasonable to expect that government funded research has become a gigantic racket, that government employment is a form of welfare, given out to people deemed worthy?  In my experience, there is much evidence to reply 'yes'.  

The FIFO would suffer cutbacks eventually, but I expect that the employees would be absorbed elsewhere in the agency.  The FAA has an Academy in Oklahoma City that is quite an impressive facility for training technicians on new equipment and training controllers.  I have spent a lot of time there.  I am confident that the FIFO pilots could be retrained or possibly transferred to other agencies.  I think that the main problem with UAV flight checks was not the fear of losing jobs.  The problem was establishing a precedent of some field engineer like me developing this solution and having it implemented.  The bureaucrats must use every need for change as a way to advance their power and prestige.  Anything that threatens this prime directive is treated like a national security threat.

The FAA had a very large number of Flight Service Stations when I started working for them.  They all disappeared in a short time because of things like ATIS.  It just made no sense to keep the FSS system operating, and a lot of people's job disappeared.  Adapting to the challenges of serving the public actually requires making such adjustments.  Technological advances erase a lot of jobs.  Why should government employees not be susceptible to this risk?

For the hardened skeptic or cynic, my account of experiences may seem unbelievable.  Maybe this author has an ax to grind with authority and the story was exaggerated to try to push some agenda.  We do need to exercise great caution because a great deal of propaganda is pushed and suspending acceptance of my story is understandable.  However, consider this.  Flight checking in FAA is a big deal.  It does not take a lot of thought to understand that having a centrally located fleet of aircraft with full time crews and a lot of sensitive electronic equipment that flies everywhere is a lot like having an air force.  It is very, very expensive, and a lot of careers are committed to keeping it going.

It is also an undeniable fact that flight checking is necessary and arduous because different organizations must coordinate their schedules and priorities to make these checks happen.  It is the only way to be certain that some equipment is working properly.  

Bureaucracies have a way of continuing their existence long after their original mission was completed.  I do not stand to gain if UAV flight checking is implemented.  Whatever intellectual property may be involved is certainly not mine.  I do feel a debt to the Anritsu Corporation, which expended considerable resources to design the device I specified to them, at no cost to the government.  Is there any personal risk for exposing corruption?  Obviously, yes, but I do not have any complaint along those lines.

Obviously, drone technology has expanded and improved.  I was considering getting a small hobby drone for fun and the available choices and capabilities are jaw-dropping.  More serious UAV are available and can obviously be specified for purposes of flight checking.  I envisioned each airport or nearby FAA Engineering Services office having a UAV that could be gas powered for extensive flight checks, or battery powered for checking NAVAIDS equipment.  This would obviate the central flight check office, except as a standards and equipment maintainer and to handle responsibilities for updating the software and corresponding FAA orders.  They might be absorbed by the FAA Technical Center.

The biggest obstacle I encountered was that FAA has no official orders to allow flight checking with UAV and it was clear that the unfortunate foot-dragging that would be hard to avoid could kill it altogether.  Trying to get government managers to work together for the common good is a pipe dream. The reality is more like getting mafia families to cooperate.  It can be done, but it must be very lucrative for all involved, and this turns my stomach.  If this is not extortion of the public, it is a good imitation.  Proud US citizens like to believe that our government is not as corrupt as the governments in countries that are constantly ridiculed for corruption.  We need to wake up.  It is worse than can be easily imagined, and accelerating downhill, in my opinion.

By the time I retired, Anritsu people were calling me to see what news I had for them.  I had none.  I was facing a long, uphill battle with no allies.  I considered the possibility of starting a company for providing flight check services for air traffic control facilities, but I am not the masochist for that job.

I think it is fair to say that UAV flight checking is a great idea for numerous reasons cited already, whether or not the reader can believe that any of my story ever happened.  Ask yourself why, given the mandate to bring UAVs into service for many good reasons, this very obvious basic first step of UAV flight checking was fought so hard.

(added 1/16/23)

The fuel savings from using flight check drones instead of FIFO aircraft, flown from central locations, does not require a detailed argument.  Just from the green standpoint, such a change is justified.

About 6 months ago, I contacted one of my former FAA engineering associates who was largely responsible for the design and implementation of the ICMS system at the Denver airport.  I asked him for contact information for the company that built what they had specified and which provided support services.  This led to a lengthy phone conversation with one of the managers of that company.  I explained my proposal for drone flight checking.  It was not hard to convince her of the many advantages of such a change and the profit potential for a contractor who was hired to develop and maintain such a program.  She told me to enjoy my retirement and forget about it once I explained to her the irrational response I had gotten from management.  She did not need to convince me that she knew very well how FAA management thinks.

I also contacted UAV enthusiasts and groups that were seeking to develop commercial applications for UAV.  I was surprised that generated exactly zero responses.  How do I interpret that total lack of interest in an obvious valuable technology, already designed, awaiting implementation, with no pesky intellectual property considerations?  Management was apparently opposed to even considering such an approach to utilizing UAVs.  It appears that people are so conditioned to responding to governmental whim as a top consideration, well above basic economics, safety, environmental protection, quality of performance and just plain sanity considerations.  If we cannot expect people to be motivated to act in the obvious interest of humanity, but to merely satisfy themselves by virtue signaling, where are we heading?


Edited 10/10/2023

I saw the name of the guy pop up on Telegram who worked at Anritsu who hinted to me that this cell coverage testing device they market might be useful for FAA flight checks.  I had forgotten about him.  I sent him a link to this blog and thanked him for giving me the idea.  I asked him if he would mind if I used his name to reference the origin.  He did not reply to that and stopped replying altogether.

He was initially responsive and friendly.  I can only imagine that being associated with some controversial technology, no matter how obviously well worth considering, and IMO, using ASAP, is not worth the risk to him.

Keep this in mind when evaluating potential value of some technological approach.  For instance, I got on a reddit site for professional chemists and asked about Randell Mills.  The non-insulting responses were from people who felt an obligation to help some silly, uninformed person avoid the pitfall of thinking Mills could be right.  This was due diligence for me.  I wanted specific empirical or theoretical reasons to disbelieve what Mills was saying.  None of those responses had any empirical or theoretical references or reasons.  I was assured by one chemist that Mills was not raising money, and therefore he was finished and his ideas therefore invalid.  I was stunned.

This was a response I might get from a stock advisor.  I wanted a chemist's reasons for rejecting Mills.

Commercial failure, in no way, demonstrates scientific invalidity.  To believe this is to accept a perfectly efficient market, a concept that has been accepted by many as real, but can hardly be considered a scientific fact.

Comments

  1. re: "Of course, cell phones companies do not probe their cellular coverage as those commercials depict. They use automated equipment like the Anritsu 2713E. "

    Well, back in the mid 90's we utilized some automated gear made by LCC Incorporated which allowed the radio/wireless engineering and optimization teams to 'plot' cell coverage for service evaluation and new-cell planning ... a device known as 'BUZZARD' (don't recall what that acronym meant) was used by corporate to evaluate actual 'call' performance in the system and for which we had dedicated personnel doing that 'drive testing' ... in the RSA (rural) cell sites it was easy to do off-the-air monitoring with a scanner that could 'step' through the assigned channel set ... and, on top of that, the marvelous Ericsson AXE212 switches spit out a voluminous data for which some analysis yielded stats that allowed problem 'traffic' areas to show up, to be posted on what we called 'the wailing wall' ... I was with ATTWS in the mid-late 90's as an RF engineer responsible for new cell engineering and initial optimization and integration into the operating system of cell sites ... there a number of parameters that can be adjusted, optimized is the word we used, to get a new cell site to 'play nice' with the cells existing around and near the new cell ... signal strength levels, hysteresis in SS, number of SS samples taken, selection of color code (in the analog cell sites) and candidate neighbor cells in the NCL (neighbor cell list) are the major parameters we could 'jockey' to adjust the traffic a cell would 'carry' ...

    ReplyDelete
  2. Interesting range of experiences you have. I didn't use the Anritsu instrument for checking cell coverage, but it can measure something that comes close to voice intelligibility, and not just carrier strength. It's been a while and I don't recall the specifics. You can probably download the manual if you're curious.

    ReplyDelete

Post a Comment

Please be polite and avoid obscenity.

Popular posts from this blog

Why Randell Mills does not appear credible to most people (yet)

Stinson Airport Small Tower Voice Switch