The direct cause of the FAA’s problems was the insertion of corrupt data during an update of the system the previous day.
FAA’s NOTAMs Debacle
On Jan. 11, I was one of the hundreds of thousands of air travelers affected by the Federal Aviation Administration’s (FAA) nearly two-hour ground stop, which led to more than 11,000 flights being cancelled or delayed. The cause was a failure in FAA’s Notice to Air Missions (NOTAM) system. The direct cause was the insertion of corrupt data during an update of the system the previous day. But the underlying cause is an obsolete and dysfunctional system that should have been rethought and replaced decades ago.
Our Notice to Air Missions system is part of an international system (called AFTN), because aviation is international. It is supposed to notify pilots, dispatchers, and others about potential safety hazards at airports and along airways relevant to a planned flight. The international version began in 1920, and the format has remained mostly unchanged since 1924. The world shifted to ASCII (upper and lower case type) in 1963, but the FAA continues to use the teletype-era all-caps format.
The major problem with NOTAMs is information overload. At any given time, FAA NOTAMs may consist of 30-to-100 pages of all-caps text, with no prioritization of what might be a serious hazard and nothing highlighted for a particular air route (e.g., Ronald Reagan Washington National Airport to Miami International). According to an aviation group called OpsGroup, the number of NOTAMs reached 500,000 in 2006 but doubled to one million NOTAMs seven years later, as FAA and other agencies continued to add notifications of things like construction cranes that are far from runways and birds congregating at or near airports. The mindset seems to be, ‘We’d better include it, in case something bad happens, and we get blamed.’
In an online aviation discussion group that I’ve been a member of for several decades, pilots and other professionals offered many critiques of NOTAMs following the ground stop. One airline pilot pointed out that the airline dispatcher’s flight plan for a specific flight extracts for the cockpit crew the departure, arrival, and alternate airports for a specific flight, and arranges the airports’ NOTAM information in flight order, but the all-caps information for each airport is a mish-mash of everything someone could think of that might be relevant, with hazard locations indicated by latitude and longitude, loads of cryptic abbreviations, and no emphasis on what might actually be important. “Hence, flight crews can spend a long time sifting through irrelevant trivia about there being a 150-foot crane a mile from the airport, or the MDA [Minimum Descent Altitude] for a particular [visual] approach on an ILS runway being adjusted from 420 ft. to 425 ft.,” the pilot wrote on the message board.
The most startling thing I found in these discussions is that the ground stop was basically unnecessary. To quote the same pilot, “Most of the active NOTAMs will have been issued days or even weeks before, so the pilots could actually have been given the previous day’s NOTAMs and just been updated with any new stuff on that day.”
In subsequent online discussions, estimates of the number of flights that could have proceeded had this decision been made ranged from 80% to 99% of that morning’s flights.
FAA’s NOTAM system is a disgrace, yet there is no announced plan to replace its obsolete computers, its ancient all-caps type, and its failure to highlight relevant safety hazards. The agency’s 2015 “FAA Resiliency Assessment Report” listed 32 air traffic control-related systems that needed change to ensure their resiliency; NOTAM was not included. In a Jan. 12 Reuters article, David Shepardson noted, “FAA has been trying to modernize the Notices to Air Mission (NOTAM) system,” but so far, the only tangible result has been changing its name to replace “Air Men” with “Air Missions.” That says something about FAA priorities.
Why Doesn’t FAA’s New Control Tower Plan Include Remote/Digital Towers?
The winter issue of Managing the Skies, the magazine of the FAA Managers Association, includes a lengthy article on the extra money Congress allocated to the Federal Aviation Administration (FAA) via the Infrastructure Investment & Jobs Act (also often referred to as the bipartisan infrastructure law). One section of the article, “Building a New Generation of Towers,” describes FAA’s plan to replace 30 smaller control towers by 2030. It discusses the agency’s recently launched Sustainable Tower Design Initiative intended to “tap innovative minds in private industry and academia…for new approaches both to design and to rapid construction.”
The article recounts an early 1960s effort that tapped major architectural firms to develop ever-grander monuments. Alas, there is not a word in this article about remote/digital towers, and this concept is also absent from FAA’s description of the program.
Remote/digital towers are certified and in operation in half a dozen European countries. They dispense with towering buildings in favor of using an array of cameras and other sensors at various locations at an airport to feed panoramic displays in a control room either on the surface or securely underground. These facilities cost a lot less to build and maintain. They also provide better performance, for example, with infrared cameras that can see approaching aircraft through low clouds, fog, and rain. They can also electronically tag aircraft viewed on the panoramic screens, and do many other things better than 20th-century “towers.”
For replacing 30 towers at smaller airports, another possibility for low-activity airports is to control several such airports from a single remote tower center (RTC). Such RTCs are certified and in operation in Germany, Norway, and Sweden and are under development in several other European countries. Whether at single airports or for groups of several, installing digital/remote towers would be faster and less costly than constructing new 20th-century towers.
Although Congress in 2018 authorized FAA to start implementing remote towers, no such projects are under way. Two state-funded remote towers have been built and are in partial operation in Leesburg, VA, and Loveland, CO. These projects began in 2015 and 2014 and have been ready for full operation for years—but are still not FAA-certified. The agency has cooperated with the project developers and has loaned some controllers, but the endless delays in certification are beyond comprehension.
Last summer, at the U.S Contract Tower Association meeting, FAA Air Traffic Organization’s (ATO) Jeffrey Vincent told attendees that “remote/digital towers are the future.” At that time, he was ATO’s Vice President for Air Traffic Services. Recently, he was shifted to being executive director of ATO’s Drone Integration Office. The “30 by 30” program would appear to be a good fit for remote/digital towers. It would be more credible if FAA finally certified the remote/digital towers at Leesburg and Loveland. And it might help if Congress, in the 2023 FAA reauthorization bill, imposed a date after which FAA could no longer build towering edifices.
Fixing the Air Traffic Organization’s Culture
The Notice to Air Missions (NOTAMs) fiasco and the Federal Aviation Administration’s (FAA) continued failure to embrace remote/digital towers are examples of a serious organizational problem. Congress created the Air Traffic Organization (ATO) in 2000, hoping that instead of being a cautious bureaucracy, it would become a “performance-based organization,” operating more like a Silicon Valley tech company to produce the world’s best, most-advanced air traffic control (ATC) system. As two decades of Government Accountability Office (GAO) and Department of Transportation (DOT) Inspector General reports have since documented, U.S. air traffic control has not been transformed. Other developed countries have pioneered remote towers, electronic flight strips, space-based ADS-B surveillance, and much more.
In 2012 the Hudson Institute commissioned me to do a peer-reviewed study of innovation within the FAA’s air traffic control system. My 54-page report was published by Hudson in Jan. 2014 (and is also available on the Reason Foundation website).
In the report, I examined seven case studies of air traffic control innovations (including controller-pilot data link, GPS landing systems, space-based ADS-B surveillance, and remote towers). In each of the seven cases, these innovations were developed and implemented sooner in other countries than in the United States (and some have still not been implemented here).
The report next offered several hypotheses to explain this difference in performance, noting that the peer countries that innovated faster and better had all separated their ATC function from their transportation agency and aviation safety regulator. The hypotheses were the following:
ATO identity as a safety agency, rather than a technology service provider. The hypothesis was that being embedded in a safety regulatory agency, rather than being regulated by it at arm’s length (as all the other aviation participants are) created an overly cautious organizational culture that is slow to implement innovation.
Loss of technical expertise. FAA engineers and software people are paid per standard federal general schedule pay categories, and work in what is, in fact, a very large bureaucracy. It is hardly surprising that many of the best and brightest can (and do) find greater satisfaction and higher pay by transitioning to private industry. This ends up putting the ATO at a disadvantage in dealing with large aerospace contractors, who sometimes design and develop more elaborate and expensive ways of meeting the ATO’s requirements.
Loss of managerial expertise. Despite Congress mandating “procurement reform,” the ATO’s procurement record features many projects that go far over budget and whose delivery extends over many years. As with engineers, the same differences in compensation and working environment lead to the best program managers being hired away by aerospace companies.
Excessive oversight. In conversations with individuals who previously served as ATO chief operating officer, I was often regaled with their frustration of having to pay attention to too many overseers: the Secretary of Transportation, the FAA Administrator, the Inspector General, GAO, Congress’s authorizing committees, Congress’s budget committees, etc. This problem applies to FAA itself as well as the ATO.
The assembled peer reviewers, all with considerable aviation and government experience, judged all four of these causes as significant, and they were generally positive about the reforms that I proposed. They were:
- Separate the ATO from FAA, putting ATO at arm’s length from the safety regulator, as is now the case in nearly all first-world countries, and has been International Civil Aviation Organization (ICAO) policy since 2001. The potential for organizational culture change would be greater if the new ATO were located somewhere other than in the FAA building—perhaps across the Potomac in a Virginia suburb.
- Shift from aviation user taxes to direct funding, similar to airports charging landing fees, rents, etc. This would be analogous to other federal entities that provide services to customers, such as the huge electric utility Tennessee Valley Authority. With its own revenue stream, the new ATO could issue long-term revenue bonds, like airports and electric utilities do, so that large capital modernizations could be financed up-front, rather than being paid for out of annual cash flow, which leads to very long periods to get improvements implemented systemwide.
- Change the governance model. Whether new ATO would be structured as a government corporation (as proposed by the Clinton administration) or a nonprofit federally chartered corporation is a decision to be debated. Both models exist in high-performance air navigation service providers overseas.
These changes are not an all-or-nothing proposition. In a 2010 article in The Journal of Air Traffic Control, former FAA Administrator Langhorne Bond and I made a stand-alone case for simply separating the ATO from FAA, making the ATO a separate federal entity, located outside Washington, DC, regulated at arm’s length by FAA (as it regulates airports, airlines, etc.). We argued that “a separate ATO would be in a much stronger position to advocate for timely implementation [of new technology] and to carry this out in a timely and cost-effective manner.”
In other words, we think separation has a good chance of leading to a more businesslike organizational culture, consistent with the new ATO becoming a high-tech service business serving aviation customers. That paper, with slight updating, was posted on the Reason Foundation website.
This is not a call to revisit air traffic control corporatization, as was debated in 2017-18 and which failed to get beyond the House Transportation & Infrastructure Committee. The coalition that backed the bill no longer exists and shows no signs of being rebuilt. Moreover, someone inserted in the huge year-end budget omnibus bill a sentence saying, “The agreement does not support any efforts to transfer the FAA’s air traffic functions to a not-for-profit, independent, private corporation.”
But as Bond and I argued in the above paper, arm’s-length separation between FAA and the ATO would remove a potential conflict of interest (self-regulation of ATC safety), be consistent with ICAO policy and global practice, and at least offer the possibility of leading to a more entrepreneurial organizational culture. That would be a meaningful reform to include in this year’s FAA reauthorization.
Originally published by Reason Foundation. Republished with permission.
For more from Budget & Tax News.
For more public policy from The Heartland Institute.