In this issue:
- The ATC “size matters” argument
- NRC committee reviewing NextGen
- New ideas for aircraft tracking
- Progress on GPS landing systems
- ERAM’s troubling outage
- News Notes
- Quotable Quotes
The “Size Matters” Argument in Air Traffic Control
In a Viewpoint column in the April 21st issue of Aviation Week, the NBAA’s Ed Bolen argued that the Nav Canada business model “isn’t scalable” because its system is only one-tenth the size of the U.S. air traffic system. And, for good measure, he repeated the point regarding the corporatized systems of Australia, New Zealand, and the U.K.
This has always been a bogus argument. If one kind of funding model-airspace users paying their air navigation service provider directly for services received, rather than paying a tax into a politicized funding allocation system-is demonstrably better and more sustainable, that model should be applicable to ANSPs over a wide range of sizes. Similarly, if one kind of governance model-a corporate form governed by a board of directors representing all the key stakeholders-is demonstrably better than a government agency accountable to numerous overseers, none of whom specializes in aviation, then that, too, should be independent of size and scale.
But these points are even more compelling when we take into account the large economies of scale that exist in the provision of air traffic management. That fact underlies the overall difference in cost-effectiveness between air traffic control in Europe and the United States, and is one of the prime driving forces for creation of a Single European Sky. The November 2013 report from Eurocontrol’s Performance Review Commission comparing ATC in Europe and the United States (2012 data) reveals the following:
| Europe | United States | |
| Geographicarea (km2) | 11.5 | 10.4 |
| ANSPs | 37 | 1 |
| En-route Centers | 63 | 20 |
| Total staffed facilities | 756 | 696 |
| Controllers | 17,700 | 15,930 |
| Total staff | 58,000 | 38,130 |
| IFR flights (M) | 9.5 | 15.2 |
| IFR flight hours (M) | 14.2 | 22.4 |
| ATC cost (B) | $11.27 | $10.45 |
From these numbers, it is easy to derive two key performance indicators. Cost per IFR flight hour averages $794 in Europe versus $467 in the United States. And annual IFR flight hours per controller are 802 in Europe versus 1406 in the United States. Clearly, the U.S. system delivers far more bang for the buck than the fragmented system in Europe. That reflects significant economies of scale.
But it gets even more interesting when we turn to the latest Global Air Navigation Services Performance Report 2013, produced by CANSO, likewise covering 2012. Although only 23 of CANSO’s 84 full members (i.e., functioning ANSPs) provided complete data, the results on the above two key performance measures, by country, are revealing. For this article, I am using data from only the six reporting ANSPs from developed countries. And due to some differences in terminology, the numbers in the CANSO report are not identical with those in the Eurocontrol report. But here are the 2012 numbers for six developed country ANSPs in the CANSO report:
| Cost/IFR flight hour | IFR flight hours/ controller | |
| Nav Canada | $339 | 1739 |
| Airways New Zealand | $431 | 697 |
| FAA ATO | $454 | 1729 |
| Nav Portugal | $565 | 1434 |
| Finavia | $702 | 620 |
| Naviair | $741 | 1033 |
As you can see, despite their smaller size (and hence less economies of scale to take advantage of), both Nav Canada and Airways NZ are delivering first-rate air traffic services at a lower cost per flight hour than the FAA’s ATO (with its large economies of scale). Nav Canada is also delivering slightly more flight hours per controller than the ATO. The other three have both much smaller airspace and lower flight activity, hence less potential economies of scale.
My take-away from this is that the cost-effectiveness and productivity of the ATO-already high-could be even higher if it were reorganized as a self-supporting ANSP, with de-politicized funding and governance.
NRC Committee Asks Probing Questions About NextGen
Over the past decade, I have discussed numerous reports about the FAA and NextGen from the Government Accountability Office and the DOT’s Office of the Inspector General. They have identified problems as well as progress, but in particular have raised important questions about fundamental issues of policy and goals that remain unresolved more than a decade after the NextGen vision became FAA policy.
In the 2012 FAA reauthorization law, Congress asked the National Research Council (of the National Academy of Sciences) to review NextGen’s enterprise architecture, as well as the safety and human factors aspects of NextGen. The resulting Committee to Review the Enterprise Architecture, Software Development Approach, and Safety and Human Factor Design of the Next Generation Air Transportation System (CREASDASHFDNGATS) released its interim report last month. Though written in bureaucratese, the interim report suggests that the Committee is asking some of the same kinds of probing questions that have been posed in recent reports by GAO and OIG.
For example, on page 2 of the Executive Summary, the report notes that the underlying NextGen vision, benefits and risks, and costs are subject to change as NextGen evolves toward implementation. Yet it expresses concern that “these changes were not fully reflected in the briefings it has received or the documents it has reviewed, and that there are not clear mechanisms to track these changes over time or to make them known to stakeholders.” Elsewhere, the interim report notes that the enterprise architecture does not “address key technical and performance parameters and relationships (including organizational and human factors considerations).”
While much of the report’s focus is on the enterprise architecture and software development and integration, I am particularly interested that the report also addresses political, economic, and cultural constraints. First, “FAA must operate within the federal political environment and under whatever financial and performance constraints and expectations are produced within it.” Also, “FAA has its own organizational culture that has developed over time,” and its “conservative safety culture can affect how quickly process and technological change can happen,” which are points that I discussed in my recent policy study, “Organization and Innovation in Air Traffic Control.” (/wp-content/uploads/2014/01/air_traffic_control_organization_innovation.pdf)
The report also expresses the Committee’s concern that it has not been able to get clear explanations from FAA management as to how it is dealing with risk and uncertainty in developing a transformation this complex-or even a “clear articulation of the significant risks perceived within the NextGen systems and software, or how the FAA would quantify or monetize these risks” to set priorities and adjust the implementation schedule. And because it has not been able, so far, to obtain a clear understanding of how the enterprise architecture is being used to make key decisions, the Committee felt compelled to ask whether instead of the architecture driving the process, “Is the architecture more of an emergent outcome from a multiplicity of separate processes?” Or in plain English, is FAA simply calling what emerges from its various efforts its “enterprise architecture”?
The Committee also raises questions about the large ongoing turnover in the controller workforce and the needed changes in skill sets and knowledge requirements for their replacements; about whether system security is being adequately addressed as more and more information is digitized and made more widely available in real time; about how unmanned aircraft will be integrated into the NextGen version of the national airspace system; and about spectrum management as newer technologies shift into different frequency bands than used by legacy systems.
What the final report, expected by mid-year, will conclude and recommend remains to be seen. But I’m impressed by the caliber of the 14 people comprising this committee-an interesting mix of software/communications experts and aviation researchers. If NextGen is to be more than just a costly hardware/software upgrade, major changes need to be made not just in the Air Traffic Organization’s risk-averse culture but in removing the politicized funding and governance that the Committee has already identified as constraints.
New Ideas for Aircraft Tracking
While we still don’t know precisely what happened to Malaysian Airlines Flight 370, the available evidence suggests that both the transponder and the ACARS were deliberately turned off, leading to the aircraft’s disappearance from air traffic control surveillance. And that raises the question of why these devices are installed in ways that permit the cockpit crew to turn them off.
The standard answer to that is fire. Generally speaking, design practice has been to provide a circuit breaker that allows the cockpit crew to disconnect electric power from any device that appears to have caught on fire or is drawing excessive current, so as to limit damage to the rest of the plane. In email correspondence following last month’s newsletter, former FAA/DOT engineer and commercial equipment developer Rick Castaldo suggested a remedy: wire such components via a “fusible link” that will self-destruct (breaking the connection) if and only if current draw gets high enough to indicate a fire in the device. Automobiles, trains, and other commercial equipment have been using such links to disable high-power devices (such as cigarette lighters, ovens, and coffee makers) for over 40 years, he pointed out.
Rick told me that after 9/11, during which the hijackers turned off the planes’ transponders, he proposed this idea within FAA, but that it went nowhere. He further noted that today, with ADS-B coming soon and with ASDE-X systems at major airports using transponder signals for surveillance and collision avoidance on the ground, there is even less justification for a shut-off provision in the cockpit, especially if fire danger can be addressed via a fusible link approach. And in terms of political support for this change, the Defense Department and Homeland Security would very likely support a change that would prevent transponders from being turned off.
Another idea has been proposed by those developing plans for the next generation of emergency locator transmitters (ELTs). These devices are designed to turn on automatically in the event of pre-defined adverse events, such as a severe G-force. The idea would be to design the new ELT to also automatically activate in the event of loss of signals from transponder, ADS-B, satellite communications, etc. The next generation ELT satellite payload, called Meosar, will be included on 72 satellites including GPS, Galileo, and Glonass. With the new ELT and Meosar in operation, the plane’s location (to within 30 meters) should be available within five minutes after the ELT is activated. Adding a “return link” to the payload would allow the Meosar rescue coordination center to remotely activate the ELT under pre-defined circumstances.
These both sound like good ideas to me.
Solid Progress on GPS Landing Systems
In a detailed article last month in Aviation Week (April 28th), John Croft provides a thorough overview of the status and progress of GBAS-the ground-based augmentation system that can and should replace legacy instrument landing systems (ILSs) as they wear out and for new runways requiring precision approach guidance.
Why is GBAS superior? First, a single system can provide precision approaches to 26 runway ends (each runway has two usable ends, depending on which way the wind is blowing), but a separate ILS is required for each runway end. Second, GBAS costs less to maintain than ILS-it has fewer components and does not require annual FAA flight calibration tests. Third, it is very precise: by augmenting the GPS signals from space with signals from four precisely located ground stations, GBAS provides vertical and horizontal accuracy of 1 meter. Fourth, GBAS facilitates close-in curved approaches and selectable glide angles that make it possible to land further down a runway when that is desirable. In addition, Croft quotes Honeywell’s Pat Reines as noting that because aircraft in the queue for takeoff don’t have to hold back from the runway end to avoid interference with ILS, each plane can save 60-90 sec. of taxi time, which can yield 4 to 5 more arrivals per day at a very busy airport.
About 3,000 aircraft have been equipped to use GBAS since FAA certification in 2006, but that capability has only been activated on 20-30% of them. Why? One major reason is that despite FAA certification, GBAS is a not-invented-here program to the FAA. Unlike with other ATC equipment (such as radars and ILSs), if an airport wants GBAS, it must buy the system itself. So far, only two U.S. airports have done so, Newark and Houston Intercontinental, both major hubs for United Continental which is a GBAS early adopter. By contrast, the ANSPs of Germany (DFS) and Australia (Airservices Australia) are embracing GBAS, with DFS having installed a Honeywell GBAS at Bremen and Airservices close to having its GBAS at Sydney certified as operational. To date, Honeywell’s SmartPath is the world’s only certified GBAS, but Indra, Park Air, Selex, and Thales are all developing such systems. DFS is testing Indra’s prototype Category III (even more precise) GBAS at busy Frankfurt airport.
With its proclaimed commitment to a revamped air traffic system based on, among other things, more-accurate surveillance and navigation using GPS, it remains a mystery why the FAA has not embraced GBAS as a key component of this transformation.
ERAM, Not a U-2, Caused Major Shutdown
Despite media reports that a high-altitude U-2 flight had “fried” the ATC system at Los Angeles Center on April 30th, the truth was more prosaic: a basic flaw in the ERAM software that is one of the building blocks for NextGen.
Though years late and at least half a billion dollars over budget, ERAM (En Route Automation Modernization) is now in continuous operation at 15 of the 20 domestic en-route Centers, including Los Angeles, with full operations at all Centers now projected by FAA for sometime next year.
What actually happened on April 30th was that an Air Force U-2 filed a high-altitude flight plan for a routine training mission that included multiple entries to and exits from airspace managed by Los Angeles Center. Despite flying at 60,000 feet-far above all normal airline and business jet travel-ERAM processed the U-2’s complex flight plan as if it was below 10,000 feet. And that meant it had to de-conflict its flight plan from numerous other aircraft flight plans, which ended up taking a huge amount of memory and interrupted monitoring of other aircraft. According to Techeye.net, controllers at LA Center were able to switch to a backup system (presumably the former legacy system) which enabled them to see planes on their screens but required non-electronic communications within the Center and to other ATC facilities. The ERAM outage was exacerbated by a simultaneous outage of the Federal Telecommunications Infrastructure (FTI).
Although FAA has announced several fixes, including changes in how altitudes are input and a larger allocation of memory, this incident raises troubling cybersecurity questions. A May 12th Reuters story quoted several white-hat hackers suggesting that the apparent lack of rigorous “what if” testing makes ERAM vulnerable to deliberate attacks. Jeff Moss of DefCon said ERAM developer Lockheed Martin should have tested for routine programming mistakes, to see what the system would do if, for example, someone input a negative altitude or a ridiculously high altitude. Dan Kaminsky of White Ops, a security firm, told Reuters that “It is certainly possible that there are other forms of flight plans that could cause similar or even worse effects.”
This bizarre event suggests to me that ERAM’s troubled history may have a few chapters yet to go.
Denmark Implementing Nationwide WAM. Danish ANSP Naviar has announced the launch of a next-generation surveillance system based on wide-angle multilateration (WAM). Provided by Comsoft, the system will include 25 sensor sites providing complete coverage of Denmark’s 23,500 sq. mi., linked to a Quadrant Central Processor. Once the system is operational, it will provide once-per-second updates of 3-D aircraft positions. The WAM system will be more accurate and cost less to maintain than the current secondary radar system. It will also be compatible with future ADS-B usage.
Inmarsat to Offer Free Global Tracking. UK satellite operator Inmarsat on May 11th announced that it will offer a new service for airliners that use its communications service. It will transmit the plane’s GPS location, heading, and speed every 15 minutes. Inmarsat estimates the cost of the new service will be about $3 million per year, which it will absorb, hoping to recoup the cost via charges for various premium services that it offers.
Alaska and Virgin America Up their RNP Efforts. Alaska Airlines says it will make routine use of Required Navigation Performance (RNP)-based approaches to Seattle-Tacoma International Airport, comprising curved approaches for landings from the north and optimized profile descents for approaches from the south. The FAA estimates that if all airlines using Sea-Tac did this, they would save 2 million gallons of jet fuel per year. Virgin America has become the first U.S. airline to achieve FAA approval for RNP 0.1 operations with its Airbus A320 aircraft. This means being able to remain within 0.1 nm of a charted course; until now, the most precise A320 capability was RNP 0.3. Five of the carrier’s A320s are equipped for RNP so far, with the other 48 to begin equipage by the end of this year. Virgin America’s 600 pilots have gone through simulator training for the new procedures.
Crimea Off-Limits to U.S. Aircraft. On April 23 the FAA issued a notice to airmen prohibiting U.S. aircraft operators from flying in Crimea and its surrounding waters. When Russia took over Crimea in March, it declared a new flight information region (FIR) overlapping the long-established Ukrainian Simferopol FIR. The FAA notice said that this presents “the potential for civil aircraft to receive confusing and conflicting air traffic control instructions” from the two different air navigation service providers, which “presents a potential hazard to civil flight operations in the disputed airspace.”
Four European ANSPs to Offer Joint Service. A new flight data processing system called Coflight is being developed jointly by the air navigation service providers of France (DSNA), Italy (ENAV), Malta (MATS), and Switzerland (Skyguide). Coflight will offer gate-to-gate flight data processing, exchange of flight plan data, and interfaces for data link communications. The current phase is defining the service capability, with the second phase developing a demonstrator, leading to feasibility trials in the third stage. Skyguide says it aims to use Coflight for its planned Virtual Center.
South African ANSP to Provide ATC Billing for Swaziland. ATNS, the ANSP of South Africa, has signed an agreement with SWACAA, the ANSP of neighboring Swaziland, to facilitate the collection of ATC overflight fees. The fees will apply to all overflights in Swazi airspace, using the ATNS surveillance network which extends as far as Mozambique and the Indian Ocean. SWACAA’s Director General said that “ATNS has a reputation of deploying the best efficient and effective aeronautical billing and collection system in the region, if not the continent.”
Cyprus to Corporatize ATC. Under an agreement with international lenders, the government of Cyprus has agreed to sell, lease, or transform a number of state-owned enterprises, including its two ports and the national lottery. The ATC system will be corporatized, converted into a self-supporting ANSP similar to those of other European countries. The change will separate the ATC service from the Civil Aviation Department by sometime in 2015.
Clarification re Iridium NEXT. In last issue’s article about options for aircraft tracking over the oceans, I mentioned the Iridium NEXT satellite constellation which will be used for the Aireon global ADS-B service. Iridium’s Matt Desch responded with the following clarification: “Airplanes do NOT need a new antenna (or electronics) once Iridium NEXT is launched. All of the current aircraft installations (and future ones) will work just fine with Iridium NEXT satellites, just like they do with current satellites. New electronics and antennas are only needed for faster connections enabled by Iridium NEXT, but there is no additional cost for airlines to just keep using their current FANS over Iridium systems.”
“Increased infrastructure spending would reduce burdens on future generations, not just by spurring growth but also by expanding the economy’s capacity and reducing deferred maintenance obligations. For example: can it be rational in the 21st century for the U.S. air traffic control system to rely on paper tracking of flight paths?”
-Lawrence Summers, Harvard University, “What the World Must Do to Kickstart Growth,” Financial Times, April 6, 2014
“If implemented properly, NextGen’s GPS backbone will allow aircraft to fly more directly and efficiently. Today, the majority of air traffic is managed through 1950s technology, resulting in longer flights, less efficient use of our air space, increased fuel consumption, and delays and congestion. The good news is that the tools and technologies exist in today’s modern jet aircraft-including the vast majority of Southwest’s 737s-to deliver immediate benefits for airlines, passengers, and the environment. While we have made some progress to implement NextGen in our nation’s ATC system, our federal government needs to do more-and do it sooner rather than later.”
-Gary Kelly, CEO, Southwest Airlines, “State of Aviation: Emerging from a Lost Decade,” LinkedIn.com, March 25, 2014
“By the early 1990s, Canada’s civil ANS [Air Navigation System] was falling behind. Skilled, dedicated people were struggling to provide an essential safety service with outdated tools and systems. Staff shortages, delays, and chronic cost overruns blocked badly needed capital upgrades. All the major stakeholders-airlines, general and business aviation, ANS employees, and the federal government-agreed that change was needed. A rare government-industry-union consensus developed around creating a not-for-profit, non-share capital corporation, with a stakeholder board of directors, to own and operate the ANS. On May 26, 1995, Nav Canada was incorporated for this purpose.”
-John Crichton, CEO, Nav Canada, “Alternative Approach,” Air Traffic Management, Issue 1 2014, pp. 16-18.