Everyday Threats to Aircraft Safety

On the morning of February 9, 1998, an American Airlines 727 on final instrument approach to Chicago’s O’Hare International Airport suddenly pitched downward. Despite the pilot’s corrective actions, the aircraft hit the ground just short of the runway. Twenty-three people were injured, and the aircraft was substantially damaged. In statements filed with the National Transportation Safety Board, American Airlines, the Allied Pilot’s Association, and the Association of Professional Flight Attendants all argued that electromagnetic interference was probably to blame for the crash. The two professional groups wrote that “circumstantial evidence” pointed to an improper glide-path signal received by the aircraft’s instrument landing system due to electromagnetic interference “from onboard electronic devices or ground-based equipment.” After investigating the accident, the safety board could find no hard evidence that such interference had occurred and blamed the accident on pilot error and an out-of-date setting of the autopilot.

In 1996, the crew of an airliner that had just taken off from Salt Lake City International Airport was informed by air-traffic control that the craft was 30 degrees off course. The pilot and copilot’s flight instruments indicated no error. On this day, the weather was good and air-traffic control was able to appropriately vector the aircraft. The flight crew later concluded that electromagnetic interference from a laptop computer being used in the first-class cabin had caused the problem. In his report on the incident, the pilot stated, “I would have really been sweating if it had been instrument flight rules in that mountain area.”

How commonly does radio frequency (RF) interference cause safety problems for commercial aircraft? We have concluded, based on several types of analyses, that RF interference from consumer electronics is unlikely to have figured in more than a few percent of commercial air accidents, if any at all, during the past 10 years. There are no documented cases of a fatal aircraft accident caused by RF interference, although it is possible that interference has been an unrecognized factor in some crashes, perhaps simply by momentarily diverting a pilot’s attention during a critical maneuver.

But there clearly is room for concern. For one thing, incidents of RF interference by passenger electronics with aircraft systems do happen regularly. We have estimated that reported events are occurring at a rate of about 15 and perhaps as many as 25 per year, based on a random sample of incident reports filed between 1995 and 2001 with the Aviation Safety Reporting System (ASRS). Additional events may go unreported. Moreover, with the rapid proliferation of consumer electronics and wireless technology, these numbers will almost certainly grow.

In addition, RF interference is a complicated and subtle phenomenon. Pilots and maintenance personnel may not readily identify problems even when they exist. Problems are likely to increase with aircraft age as avionics are removed, modified, or changed and as gaskets, shielding, and grounding straps become corroded, get left out, or are otherwise rendered ineffective. In newer “fly-by-wire” aircraft, which depend more heavily on complex electronic circuitry, the number of system locations in which RF interference might cause difficulties grows. Aircraft designers and manufactures, as well as airline maintenance personnel, are aware of these problems, and they work hard to minimize vulnerabilities. But faced with the slimming down of work forces, expanding job responsibilities, and the retirement of older personnel who had specialized knowledge and experience in electromagnetic compatibility, the potential for problems increases.

The electromagnetic environment

A commercial airliner is a study in electronic complexity. Aircraft systems–marker beacons, distance-measuring equipment, traffic-alert and collision-avoidance systems, microwave-landing systems, and Global Positioning Systems, among many others–operate across a wide range of radio frequencies. Likewise, a host of consumer electronic devices–laptops, cell phones, game systems, CD players, and the like–produce emissions that range across all of these frequencies. In addition to producing emissions at their nominal design frequencies, passenger electronic devices often also produce emissions at other frequencies due to harmonics and other mechanisms. The physical environment of the aircraft further complicates matters. Aircraft cabins are large metallic tubes that can act as resonant cavities at some frequencies, and an aircraft’s windows, which are basically openings in a conducting plane, can radiate as slot antennas.

New consumer electronic devices are required to meet RF emission standards set by the Federal Communications Commission (FCC). These standards are intended to prevent interference with other users of the electromagnetic spectrum. However, because of problems in testing, enforcement, and other areas, it is not clear that all new devices conform to these standards. In addition, as devices get banged around in use and get modified or repaired by owners or service personnel with limited training, some of them can begin to produce emissions that exceed their designed specifications.

RF interference from consumer electronics can affect various flight-critical aircraft electronic systems in several ways. Radiation can pass out of the windows and enter antennas located on the outside of the aircraft. Emissions also can directly enter electronic devices inside the aircraft or induce currents in wiring that is connected to those devices. Of course, aircraft systems have been designed and tested to minimize the risks of such interference. But if emissions from passenger electronics are great enough, these measures may be insufficient to protect some properly operating systems. Of greater concern is the possibility that some of an aircraft’s systems may not be operating as precisely as intended, perhaps due to aging or maintenance. In addition, devices that individually test as reliable may exhibit problems when operated together. RF interference with aircraft systems can often involve very subtle interactions among a number of different system components, and such interference may be difficult or impossible to detect through routine testing and preflight checkout procedures.

Intentional transmitters in the hands of passengers further complicate the picture. As all Americans now know as a result of the heart-wrenching calls made from Flight 93 on Sept. 11, 2001, cell phones can be operated successfully from the air. The FCC bans cell phones from use in flight, not primarily for safety reasons but to reduce their interference with ground-based communications systems. At altitude, a cell phone illuminates many base stations, and their ban is intended to avoid overloading the system that coordinates the hand-off of calls from one base station to the next. Clearly, this ban also helps protect aircraft from on-board sources of electromagnetic radiation–but it is not a perfect solution. Some passengers occasionally turn on their cell phones during a flight, as evidenced by incident reports to the ASRS and a small survey we conducted of passengers, not to mention by numerous informal observations. In addition, some passengers forget to turn off their phones when flying. Such phones emit periodic “here I am, are you out there?” signals in order to locate and coordinate with base stations. Modern phones may first try this at low power, but if they do not receive a response, they try again at higher power. How common such short transmissions are on today’s airliners is not clear. There are probably a few such phones occasionally transmitting on many flights.

In-flight use of cell phones also is likely to increase under pressure from passengers who want to stay connected. Several airlines are now installing “microcell” systems in their cabins that will support cell phone use during flight. These systems enable passengers’ phones to communicate at low power with an on-board base station, which in turn is connected, via a dedicated radio link, to the terrestrial telephone system. Many airlines also are moving to install electrical power and Ethernet connections at passengers’ seats. Because of the high data rates involved in computer processors and in Ethernet communications, and because the systems are being connected directly to aircraft electrical systems, these in-seat facilities hold some potential to cause RF interference with aircraft systems.

These are not the only sources of radiation likely to appear in aircraft cabins. Wireless systems are growing in popularity. Soon, customers will want to operate laptops, personal digital assistants, or game systems in different parts of the cabin while networked together using wireless technology. For example, the sales representative in seat 12C, preparing for an upcoming presentation to a potential customer, may want to use his wireless system to ask an engineer in seat 19F for help with some technical details on their product. Similarly, teenagers in 15B, 16C, and 23F may want to challenge each other in the latest multiperson computer game.

Pressures on airlines will probably also grow to support wireless Internet connections in aircraft cabins. In a few more years, even more potentially problematic sources of radiation may become an issue. For example, “ultrawideband” emissions have been demonstrated to interfere with the operation of aircraft Global Positioning Systems, and GPS is playing an increasingly central role in plans for future navigation and instrument approach systems.

Managing hazards

Safety purists might argue that airlines should simply ban the use of all consumer electronic devices in aircraft cabins under the authority they already have through existing Federal Aviation Administration (FAA) regulations. The FAA specifies that, “no person may operate…any portable electronic device on any…aircraft” unless an airline has determined that use of the device “will not cause interference with the navigation or communication system of the aircraft on which it is to be used.”

It is unlikely, however, that airlines will issue such a ban. Competitive pressures among airlines are large and growing. Business travelers, who are the most likely to want to stay connected and networked, are also the most profitable group of customers. There will be enormous pressure to introduce new services as airlines search for sources of comparative advantage. As long as one major airline allows or supports a service, there will be pressure on others to do the same. Further, since some of these technologies carry clear productivity and other benefits, it would be inappropriate to restrict their use through an overly precautionary policy if more balanced risk management solutions could be developed.

Instead, there are a number of management and control actions that parties on all sides of the issue can take to help improve air safety. For one thing, airlines, aircraft and equipment manufactures, and regulators should make greater use of the classic tools of risk analysis to examine the problem of RF interference. However, because of the enormous diversity and complexity of the systems involved, the constantly changing aircraft environment, and the limited analytical resources, we believe that such conventional studies will not be sufficient for identifying and assessing all important potential accident sequences in a timely manor.

Greater progress can be made through five broad strategies that will foster adaptive management and control. These strategies involve:

• Paying careful attention to aircraft equipment design and certification, and to quality control in maintenance. Obviously, airlines should maintain due vigilance regarding their existing equipment and systems. Moreover, airlines should move with great caution as they proceed to consider new aircraft systems. The potential for problems associated with emerging wireless systems is probably large. Since individual airlines may not have the resources to adequately evaluate all systems under development, a joint effort is indicated, and in the interests of public safety, some federal money should be provided to augment airline resources. FAA budgets have long been tight, and today they are stretched even thinner by the demands imposed in the aftermath of the 2001 terrorist attacks. The FAA, the FCC, airlines, and equipment manufacturers should form a joint industry-government cooperative program to perform evaluation and testing, and Congress should appropriate funds to support the federal contribution to this undertaking. In addition to expanding the extent and quality of analysis and testing, such a program also would help to reduce redundant testing efforts across the industry. And since participation would be mandatory for all airlines, it would improve information sharing and eliminate free riding. Today, because of competitive considerations, airlines that have invested heavily in interference testing are sometimes understandably reluctant to share results with other lines that have invested less heavily.
• Augmenting the Aviation Safety Reporting System so that once again it can support statistically meaningful analysis of events and trends. The ASRS is operated by the National Aeronautics and Space Administration (NASA) in cooperation with the FAA. Flight crews, controllers, maintenance staff, and others send confidential reports concerning observed safety problems to NASA, where the reports are summarized in a form that assures confidentiality and avoids punitive consequences for those reporting. The system has become a cornerstone of aviation safety, as well as a model for other fields, such as medicine. The ASRS has received more than 500,000 incident reports and issued more than 4,000 safety alerts, and outside researchers have drawn on the database to produce at least 60 reports and papers. But NASA now can afford to enter only 15 percent to 20 percent of the incident reports it receives. Entries are chosen on the basis of a “watch list.” Between 1995 and 2001, only 10 percent of the reports, about half of all entries being made, were randomly selected. Because of budget cuts, the practice of including an identifiable random sample has been dropped. Thus, ASRS can no longer be used to do full, statistically valid time series studies of all types of incidents, including those involving RF interference. Clearly, Congress should provide budgetary support to reinstate the random sample.
• Improving characterization and analysis of the RF environment on aircraft. The only reported studies of potential interference from consumer electronic devices have involved static tests conducted on the ground. With FAA support the authors are undertaking a pilot program of in-flight measurements (as well as measurements of a sample of the public’s laptops and game systems), but it will be some time before results are available. Such measurements could be made a routine function on all flights. Modern flight data recorders–the familiar “black boxes” that serve as tools for investigating aircraft crashes–have hundreds of channels for recording data. Major airlines now routinely apply data-mining methods to the records from each flight in order to improve operational efficiency and quality assurance and to search for anomalies that may be indicative of problems. It would be relatively straightforward to install in aircraft cabins a set of RF detectors that would continuously monitor field strength in several spectral bands and record the data in the black box. Analysts could then include an examination of the cabin electromagnetic environment in their search for potential interference problems.
• Developing and deploying simple real-time tools to help flight crews detect RF emissions. If airline cabins were equipped with RF detectors, then flight crews could take corrective action when strong electromagnetic emissions occurred. The utility of equipping flight crews with easy-to-use hand-held RF detectors also warrants investigation. If such observations ultimately identify particular types of electronic devices that are seriously troublesome, then legal or other mechanisms should be available to keep them off of airliners in the future. Currently, there is no systematic way to keep offending devices off of flights.
• Paying greater attention to managing the RF emissions of consumer electronics. The FCC currently does not confer with the FAA when establishing RF emission standards for consumer devices. Such coordination would be desirable. In addition, the national debate over the management of the electromagnetic spectrum and wireless technology should pay greater attention to the consequences that different policies will have for the aircraft environment. If the expected growth of wireless technology leads to interference problems that are sufficiently grave, then it may prove necessary to adopt more aggressive control measures. For example, the FCC might require manufacturers to include override capability in wireless devices so that they could be turned off by a centrally transmitted control signal during critical phases of flight, such as take off and final approach. Such a “silencing” capability might also prove beneficial in other critical settings, such as hospital critical care facilities, as well as in such social settings as theaters, restaurants, and library reading rooms. This type of regulation, however, would raise important questions of civil liberties, social vulnerability, and the potential for “common-mode failure” in important communications systems, and such a requirement should not be imposed without careful analysis and a balancing of risks, costs, and benefits.

Taken together, these actions will enable regulators and the airline industry to better characterize and adaptively manage the risk that RF interference from consumer electronics poses to aviation safety. In an industry that has eliminated or is effectively managing most large and obvious sources of risk, such persistent risks increasingly warrant attention.