No other topic has produced the same level of reader e-mail and feedback as that of personal electronic devices (PEDs) aboard aircraft. The potential and recorded effects of these devices on aircraft systems has clearly been an issue for many readers, and I think recounting some of those experiences and comments is worthwhile. Please understand that I am summarizing many messages, so any errors caused by that process are mine, as well as any generalizations. Very few people were uncommitted in their comments.
I heard from several labs and consulting engineering groups, both civil and military, who have been wrestling with this problem. Their general experiences have been parallel, in that they have had difficulty reproducing reported flight events. But they diverge as to conclusions.
The military testing indicates that errors of navigation could not be caused by any PED they have seen so far. Meanwhile, civil tests support the "denial of service" model I reported earlier, where a system might become unavailable, rather than give misleading navigation.
Both groups were frustrated by their inability to reproduce some situations that were reported by credible flight crews. This reinforces the impression that PEDS alone did not cause these situations; rather multiple events and a specific mode of operation triggered combinatorial situations.
Many test engineers also sent information about known interference situations they encountered, ranging from video games installed in seat-backs affecting on-board radio systems to cell phones triggering system malfunctions on-board aircraft in flight. These situations were repeatable and identifiable, and raised concerns in their minds as to how aircraft safety might be affected.
One letter from Daniel Hawkes, head of Avionics Systems, UK Civil Aviation Authority (email@example.com) was particularly important, because he was able to provide some excellent supporting analysis with the fault report as follows:
"The UK CAA has an advantage over other authorities in that, by law, the industry is required to report events which affect safety to CAA. As a result, we get regular reports of interference effects due to passenger carry-on equipment, and we have seen an eight-fold increase in cell phone interference reports during 1999. This is consistent with the massive growth in cell phone ownership. The most recent report involved an incident in December 1999 where the pressurization outflows were opened in a 747-400 causing the cabin altitude to climb when a passenger on the upper deck used his Iridium satphone. (Similar reports from hangar maintenance staff have been made over the years where walkie talkies have had the same effect.)
One problem in associating cell phone use to aircraft effects and trying to repeat the event is the sophistication of the digital phone itself. The output power varies depending on the traffic on the network and the quality of the link. On the ground, where a good link is obtained, the power falls to minimum (possibly a few hundred milliwatts), whereas in the air, where the link is weak due to distance, the power increases to max (possibly 3 watts). This factor together with all the other variables of the flight situation make repeatability elusive. Also, you can never be sure which passenger’s phone was actually causing the problem.
The fact that flight crews regularly acknowledge hearing the distinctive blip blip blippity blip cell phone registration transmission in their headphones confirms that the interference does penetrate the systems. The question is, can you be sure that an adverse effect will not arise? In our view, No."
In another letter, an airline project engineer made this excellent observation about my (now admittedly incorrect) comment that I could see no problem with use of cell phones prior to flight in the aircraft:
"There are many phases of ground operations, including prior to departure taxi operations, where aircraft systems may be affected. For example, navigation systems might be updating the flight management system, or voice or data communications might be corrupted. Miscommunications or navigation errors experienced prior to takeoff may represent a flight hazard. If operators want to allow the use of PEDs at any time while on-board aircraft, including cell phone use during ground operations, then they must be sure that no aircraft systems will be affected by their use."
He is quite right, once on-board the aircraft, all of these transmitters need to be off, or it must be known that they are of no hazard. In my desire to accommodate passenger concerns, this had seemed like a good and workable compromise, but obviously, I wasn’t looking hard enough.
Several people have suggested that a way to reduce "furtive" cell phone use is for airlines to allow some free calls for passengers on those planes equipped with set-back phones. This has some appeal, as many passenger and industry groups who oppose the ban on cell phones in aircraft do so because they feel it is simply a ploy by airlines to force them to use expensive on-board phone systems. Some on-board phone access for passengers would clearly defuse this argument, and greatly encourage compliance.
One other area that was well addressed by Daniel Hawkes was the propagation mechanism into "approved systems," and he provided a CAA report on cell phone use in aircraft. It has the following very useful section on existing industry test limits, which shows the gradual shift in compliance limits and how a ship can have approved systems of differing compliance:
1) Aircraft equipment has to be approved by the responsible authority in the country of manufacture and installed in the aircraft in an approved manner. An internationally agreed aviation standard exists for qualifying aircraft equipment for approval with respect to the extremes of its operating environment, including exposure to interference. The standard, known in Europe as EUROCAE ED-14 and in the United States as RTCA DO-160, defines test procedures, which demonstrate equipment’s capability to withstand extremes of temperature, vibration, etc. For radiated and conducted radio frequency susceptibility, tests are defined to determine whether the equipment will accept a level of interference coupled into the equipment by a radiated field or by direct injection into the system through its connecting wiring. The severity of the tests depends on the equipment’s criticality, as defined in the standard.
2) The standard has been progressively updated, with the qualification tests becoming generally more severe. The intensity of the radiated interference signal to be used in the test for the most critical category of equipment (at frequencies generally applicable to mobile telephones) is expressed in units of volts per meter.
3) Thus, equipment approved prior to 1984 would have been tested in accordance with either ED-14/DO-160 or ED-14A/DO160A. Similarly, ED-14B/DO-160B would have been used up to 1989. Equipment approved to these standards can remain in production and continue to be installed in newly built aircraft, which are derivatives of types first certificated in the same period.
4) The equipment in an aircraft may include items qualified to ED-14/DO-160 (pre-1980), ED-14A/DO160A (pre-1984), and ED-14B/DO-160B (pre-1989). The significance of this point is that the installed equipment is a mix of items with differing interference susceptibility.
As I mentioned earlier in this series, it is possible to have a compliant product, which has relatively weak susceptibility testing limits, making it potentially vulnerable to PED emissions. This chart helps to quantify those incremental approval limits and illustrates how a system could be affected. The CAA report goes on to do the following calculations to show the potential for interaction and susceptibility:
5) Mobile telephones (excluding satellite telephones) operate in the frequency bands of approximately 415 MHz, 900 MHz or 1800 MHz. (Some regions of the world use slightly different frequencies). For the digital telephones, the transmitter maximum output power typically ranges from 1 to 3 watts. Transmission can occur in the "Standby" and "Call" operating modes. Telephones with an alarm function may switch automatically to "On" from the "Off" condition when the alarm activates.
6) Applying fundamental principles, the intensity E (known as field strength) of the transmission at a distance D from a telephone transmitting P watts of radio frequency power in a free, unobstructed space, can be estimated using the formula: E = (7 ÖP) divided by the distance D. Thus, for a 2-watt telephone, the field strength in free space at 1 meter distance from the telephone is approximately 10 volts per meter, and at 100 meters distance, approximately 0.1 volt per meter.
7) However, in the confines of a metallic aircraft fuselage, complex propagation paths arise due to reflections from the metallic structure, which can lead to signal cancellation or reinforcement at different locations in the aircraft. Although the free space equation does not give reliable results under these conditions, it is reasonable to conclude that the field strength of the interfering telephone transmission will exceed by a significant margin the levels used in susceptibility tests for critical avionic equipment qualified prior to 1989.
Following a study of the problem, an industry committee sponsored by RTCA Inc., a body acting as a federal advisory agency of the U.S. government, indicated in its report DO-233 of August 1996, that at interference levels of this magnitude, it is impossible to guarantee that aircraft electronic systems would not be disturbed. The report went on to cite 100 volts per meter as being an equipment qualification level for which harmful interference from transmitters within the cabin would not arise. As an example of the problem, an instrument approach and landing guidance radio installed in a derivative model aircraft built in 1999 may be qualified in accordance with the earlier DO-160 standard having been tested only to a level to 0.1 volt per meter up to 1215 MHz. This test would not have checked its susceptibility at an interference frequency and field strength equivalent to that transmitted by a 2-watt mobile telephone operating at 1800 MHz.
With the imminent appearance of more "wireless" systems of all kinds, one should keep in mind that the composite picture presented by a modern aircraft is one of potential mixed susceptibility, coupled with considerable difficulty in demonstrating a conclusive "safe or susceptible" status. This is due to the complex interaction of systems and potentially offending PEDs. For this reason, I believe that a workable approach to safety in this area has to restrict the use of PEDs, and allow for creation and certification of "non-offending" very low emission systems, and permit no active transmitters.
Only consistent communication with passengers on this issue can achieve compliance. Having one or more airlines allow use, while others do not, is not going to get us to a safe conclusion.
In addition, some kind of simple RF monitor on-board aircraft should be considered (either fixed installation or hand-held) to allow quick identification of potentially offending systems, as well as access to on-board phones to reduce the likelihood of "furtive use" in flight.
It is critical to understand that it is not a requirement for PED emissions to be specifically "in-band" to cause system problems aboard an aircraft. Many public and industry groups pose this "proof of no hazard" argument to stop restriction of cell phone use. If energy levels are high enough, the problem is really one of energy susceptibility, and may affect any system on-board the aircraft, not just a radio per se.
In addition, many possible situations exist where PED emissions may have either in-band or mixer/image components that combine to stop the use of an on-board radio system, and large levels are not required if a propagation path to the aircraft antenna is possible.
Many people remain skeptical of incidents reported by pilots who have proof of PED interference. So, I am indebted to one RTCA committee member for providing the following link. It covers a rotating incident summary from NASA, which will rivet the attention of anyone who reads it: http://olias.arc.nasa.gov/asrs/repsets.htm. (Select the PED section towards the bottom of the page for the last 50 reported events.)
The high number of incidents which include compass error suggests strongly to me that flux valve signals (which are very low-level analog lines, plus the flux valve itself is intrinsically unshielded in order to operate) are being readily influenced by on-board PED transmissions whether unintentional or deliberate. This explains the repeated compass/nav error reports, which suddenly correct themselves after a cabin announcement to stop use of phones and other items. Note carefully the reports that also identify CD players and games. Careful flight crew experiments have been conducted in-flight to prove the relationship of cause and effect.
The London Sunday Times also has a very disturbing article at: www.sunday-times.co.uk/news/pages/tim/99/08/06/timfgnfar01001.html?19 further detailing a recent incident with identical causes and system interaction.
Both experimental and flight report evidence clearly supports some swift action to control and qualify on-board PEDs. Continued inaction in this area can only lead to fatal incidents in the future.
Walter Shawee welcomes reader comments and may be reached by e-mail at firstname.lastname@example.org