Cameras Being Deployed on Aircraft to Improve Safety, Security

Aviation industry lags rail, maritime sectors in use of video recordings
The cameras are coming. In the not-too-distant future transport category aircraft might well be equipped with closed circuit television systems to cover both interior spaces and the exterior of the aircraft as the technology to do so becomes more compact and capable.

The coming changes import added costs to equip airplanes and significant changes in operational procedures. The net impact could improve both safety and security. The legal ramifications are not yet fully resolved but in-flight video recordings could reduce an operator's liability exposure, and in fact camera coverage may become a necessary adjunct of an operator's legal protection. Installation and maintenance costs of video imagery remain uncertain, but government funding for installation may be needed given the dire financial straits of the industry.

At a Society of Automotive Engineers (SAE) transportation recorder symposium last week, a new acronym was presented: AIR. The letters stand for airborne image recorder. In the next few years this acronym could take its place alongside the flight data and cockpit voice recorders (FDR/CVR) in sort of a trilogy of recorder capability: FDR/CVR/AIR.

The National Transportation Safety Board (NTSB) co-sponsored the seminar as a follow-up to the board's recorder symposium four years ago (see ASW, May 10, 1999).

The need for a video recording in the cockpit has its roots in the advance of aviation technology. The advent of data-links and text messages, developed in part to reduce the volume of voice message traffic on the radio, meant that the traditional CVR would not capture key elements of air-ground communications. Moreover, the displays in modern glass cockpits are graphical interpretations of the instruments they have replaced, and any anomalies in the translation of electronic data to the visual presentation might present misleading information.

Image recording was seen as a means of picking up CNS/ATM (communication, navigation and surveillance/air traffic management) information as received and displayed to crews in the cockpit. Moreover, image recording of the cockpit was seen as complementing the FDR/CVR, providing a better sense of the environment (e.g., smoke), non-verbal crew communications, a better appreciation for the crew's workload, and so forth.

Accident investigators have asked for a cockpit video recording. Notably, in April 2000 the NTSB called for a cockpit video recording capability in the wake of its investigation into the 1999 crash of EgyptAir Flight 990, and the lack of cockpit imagery in this and other cases that complicated the investigation (see ASW, April 17, 2000 and Nov. 6, 2000). In addition, the Transportation Safety Board (TSB) of Canada has also called for cockpit video recording as a result of its investigation into the 1998 crash of Swissair Flight 111 (see ASW, April 7).

From the advance of CNS/ATM, the limitations of FDR/CVR technology, and the need to better understand the human/machine interface in incidents and accidents, the imperative has emerged for video recording. Indeed, over the last seven years, a working group formed under the auspices of the European Organization for Civil Aviation Equipment (EUROCAE) has been meeting, and in March produced a minimum operational performance specification (MOPS) that sets forth performance parameters for image recording in the cockpit. The document (EUROCAE number ED-112) does not spell out the number of cameras to be installed; rather, it sets forth the desired performance, in terms of update rate (4-5 frames per second) and resolution (sufficient to read the instrument panel).

Over the next three to four years, EUROCAE's ED-112 specification will be transformed into a Federal Aviation Administration (FAA) technical standard order (TSO). In terms of regulatory activity, the train has left the station, and operators can expect to see a specific requirement, with implementation deadlines, for cockpit video recording.

Two other factors have accelerated the push for video. The terrorist attacks of Sept. 11, 2001, led to the installation of hardened cockpit doors, a massive retrofit accomplished by the FAA's April 9 deadline. However, the area immediately outside the cockpit will require camera surveillance, with a monitor in the cockpit. Pilots have to control access to the cockpit; they have to ensure that anyone seeking access to the cockpit does not pose a threat. Moreover, they must be able to do so from their normal seated position, hence the need for video cameras. As an example, by November 2003, all UK-registered airplanes are required to have live image viewing from the pilots' normal seated position of the area directly outside the cockpit door. The FAA and Europe's Joint Aviation Authorities (JAA) have the same regulation under consideration on a longer time scale.

The live imagery, of course, can be recorded to document any future terrorist attacks on the cockpit. By the way, no performance specification has been developed to date for capturing imagery outside of the cockpit. Guidelines are needed urgently, according to symposium participants.

The need for expanded coverage of the entire cabin also is pushing the technology. Mike Horne, managing director of AD Aerospace, a UK manufacturer of video surveillance equipment, said, "An attack won't start just outside the cockpit."

Moreover, he pointed to the rise in disruptive passenger incidents, in some cases involving outright attacks, as adding to the need for complete video coverage of the cabin. Horne pointed out that the FAA investigated 314 incidents in 2002, and three Japanese airlines surveyed that same year posted 570 such incidents. These numbers show that the problem may be vastly under-reported. It is "hard to believe," Horne said, that Japan with its smaller airline industry would have nearly twice as many incidents. With pilots locked in the cockpit, the flight attendants now must face these incidents by themselves. Horne cited a number of "potentially dangerous and certainly costly incidents" where a video recording would have been extremely useful in subsequent court proceedings.

Court proceedings often devolve into the flight attendant's word against the alleged perpetrators. A video recording of the event, Horne pointed out, would help to successfully prosecute these cases.

Technical challenges
"Putting video cameras on aircraft is not a trivial thing," Horne said. As in the case of FDR/CVR boxes, an AIR would have to be hardened against shock, heat, water penetration and such. As in the case of its FDR/CVR cousins, the AIR would need a recorder independent power supply (RIPS) to guard against gaps in the data The RIPS would have to kick in within 50 milliseconds of the aircraft losing power, said Jim Elliott, the marketing manager for Michigan-based Smiths Aerospace Electronic Systems. Elliott said it is especially important to "maintain data collection during a very critical period" when electrical and other systems may be failing.

Horne ticked off some of the items affecting installation and performance:

The light range: Even within a single picture, it can vary by a factor of 100,000, between the brightest scenes above the clouds to a dimmed passenger cabin.
The temperature can vary from -140� F on the aircraft's exterior to more than 120� F inside an aircraft parked in the desert.
The power supply can vary and is subject to dropouts during engine start.
The cabin pressure changes, and the rate of change can be rapid should decompression occur.
High reliability and ease of maintainability must be considered.
Picture quality is a big issue. To record the instrument panel, a high-resolution camera operating at a low frame rate (1/sec.) is preferred, but that capability requires a great deal of memory. For recording human and other activity, a lower resolution and 4-5 frames per second is preferable, as the imagery eats up less storage. "You almost need two separate cameras" to cover the cockpit, Horne said. His company is experimenting with a variety of possible cockpit camera installations to arrive at the best trade-off between resolution and frame rate.

There also is a trade-off between field of view and resolution. The wider the area covered by the camera, the fewer the number of cameras that might be needed, but at a "cost" of lower resolution.

A color capability also requires more memory. With black and white, Horne explained, "You lose the color but it's more sensitive, so you get higher resolution and it works better in low light."

However, to best capture multicolored glass cockpit displays, a color recording would be essential.

For cabin coverage, a color capability would not be critical but nonetheless might be preferred. "We live in a color world. People like it better," said Mike Zwolinski of RemoteReality, a Massachusetts-based video surveillance manufacturer.

Digital or analog? With digital processing (not possible with analog technology), Zwolinski said "every pixel can be enhanced to bring out the detail."

The availability of wide-area digital cameras makes it possible to cover the cabin of an airliner with just four cameras. The pictures can be fed to a display in the cockpit, and selected imagery could even be transmitted to an air marshal on board, who would receive the camera coverage on his PDA (personal digital assistant). With satellite linkage, imagery could be transmitted to the ground, Zwolinski added.

The technology is at hand, Zwolinski said, to cover the entire cabin with just four cameras. The installations would be very unobtrusive, with two in the ceiling and the other two mounted just below the overhead bins on the side of the cabin.

Catching up
The aviation industry is in some respects behind the video technology deployed in other transportation modes. Hundreds of black and white video recorders have been installed in railroad locomotives. Aimed down the center of the track and focused about 100 ft. ahead, these cameras (which capture sound) record the entire trip a train takes. They are particularly useful for recording incidents and accidents at highway/rail crossings. They also have been useful in recording what were euphemistically described as "trespasser suicides."

The LocoCAM manufactured by GE Transportation Systems costs about $21,000 to install on a locomotive, but the average settlement cost from highway crossing and trespassing accidents amounts to some $3.2 million per case, according to GE's Joe Aguayo.

On LocoCAM, the frame rate increases with the locomotive's speed. Should an accident occur, the last few minutes of video can be transmitted from the locomotive via satellite to the company, Aguaro added.

Capturing the sound of the horn along with the imagery "is vital if we go to litigation," said Gregory Lowe of Norfolk Southern railway.

Lowe showed the video of a semitrailer truck pulling onto the crossing just as the train was approaching, with the engineer sounding the horn. The locomotive struck right at the middle of the trailer. Later, the driver claimed the train "came out of nowhere" and no horn was heard. When the plaintiff's lawyer viewed the video, he dropped the case. The video included the train's speed (26 mph.) and a red box showing the precise seconds during which the horn was sounding.

Of 100 incidents over a two-year period, Lowe said, 98 of the cases were dropped "thanks to the video evidence."

Tom King, of Northrop Grumman's Marine Information Systems, said video technology is being deployed on ships. The imagery on the ship's navigation radar is recorded, and CCTV cameras capture activity on the bridge, to include the bridge wings. In addition, he said, CCTV coverage is being expanded to cover internal spaces and external areas of the ship where cargo containers are stacked. The need for improved security has provided the impetus for these developments.

By the way, an increasing number of U.S. school buses are being fitted with video recorders. According to Steve Belden, president of Texas-based Zepco, a manufacturer of vehicle recorders, a school bus might feature cameras to cover the driver, the step-up entranceway, and students sitting in rows behind the driver. A supplementary camera might be placed in the back, to cover activity in the last rows. In addition, cameras can be installed to cover the flip-out passing barrier mounted on the front fender. In this respect, the video coverage for school buses goes well beyond the debate over cockpit video in the aviation industry.

Privacy protection
Can video recordings of accidents find their way into the public domain? The final moments of aircrews might become the voyeuristic grist of "snuff films" on the Internet. In addition, employees may find the constant camera coverage intrusive.

Christopher Julius of the NTSB's office of general counsel said, "There is a lot of emotion and a lot of concern about privacy issues."

In the railroad industry, video may be expanded to cover activity inside the locomotive's control cabin. That initiative is a close analog to video coverage in the cockpit of an airliner.

Robert Harvey of the Brotherhood of Locomotive Engineers, said crewmembers occupy the cab of a locomotive for up to 12 hours per day. Much of the time the locomotive is not moving and activity during this period is seldom safety critical. He suggested that the video recordings should only be used for accident investigations, that the field of view should be restricted to the engineer's look outside the locomotive and the hand movement of controls, and that an erase feature should be incorporated, which could be activated by the crew at the end of a duty tour, "provided no FRA [Federal Railway Administration] or NTSB reportable accident has occurred."

The erase feature could be sort of a "dual key" system, where the crew would receive by radio an authorization code from the company to erase the information.

"The government is asking for this [video] technology, so I think the government has some responsibility to assure that the data is protected," Harvey declared.

Capt. Terry McVenes from the Air Line Pilots Association (ALPA) offered this perspective:

"Without a doubt, protecting recorded data in any form on a global basis is a monumental task. Therefore if we are to go down the road of cockpit imagery, the most reasonable solution from the ALPA perspective is to limit the field of view of video recording devices. By focusing the cameras only on such things as cockpit instruments, displays and annunciators, such image data might aid in an accident investigation if it is used in conjunction with other investigative techniques.

"However, it must be kept in mind that recorded data, in whatever form used, is just one piece of the accident puzzle. We should never let image data be used as an investigative shortcut that could prevent the accurate identification of all safety deficiencies that may have contributed to the accident.

"ALPA believes that we must first develop and implement the best possible protective measures of recorded data prior to the installation of any image recording devices in the cockpit. As a starting point, we feel that the following constraints should be placed on sensitive recorded information:

1. Only the accident investigation authorities should be legally or technically able to conduct readouts of the information. This should be technically feasible via data encryption methods.

2. The recorded information should never be released to any parties outside of the investigation.

3. A narrative summary of the recorded images could be produced but not released to any parties outside of the investigation.

4. Tort litigation legislation needs to be strengthened and extended to address criminal proceedings as to how CVR and image data information can be used. Similar legal protections should be in place in all ICAO [International Civil Aviation Organization] signatories.

5. At the conclusion of the investigation, the recorded image information should be either destroyed or permanently retained only by the accident investigation authority."

Complete coverage
The signs point to the much broader use of CCTV than to cover the cockpit and the entrance area immediately behind the cockpit bulkhead. Someday, the typical airliner may be outfitted with a dozen or more cameras. Consider:

JetBlue Airways [JBLU] already is equipping its aircraft with three cameras to cover the area just behind the cockpit. One camera looks forward to cover the cockpit door; a second covers the entranceway on the left; a third covers the service area on the right.
Full cabin coverage with 4-5 cameras for a narrowbody, and 8-10 for widebody, Horne said.
Two cameras covering the fore and aft belly holds.
Swissair's Modification Plus program, in which fire hardened infrared (IR) cameras were installed to cover the attic space over the forward cabin and the electronics and equipment (E&E) bay (see ASW, July 30, 2001).
Adding a belly camera with a 360� field of view. During ground operations, this camera could provide a security function, enabling pilots in the cockpit (or receivers at another location) to monitor fueling, loading, and so forth. In flight, the camera would record flap and landing gear functioning, thrust reverser operation, and might also record tailstrikes.
Adding a camera topside on the fuselage or tailfin to record rudder, elevator, aileron and spoiler movement, among other systems. The functioning of flight control surfaces on the main wing often cannot be seen from the cockpit.
The total comes to some 14 separate cameras for a widebody, perhaps as many as 16-18 to cover inaccessible areas, wheel wells, and other areas. Swissair's installation allows for pilots to select the view from each camera, and this feature would have to be integrated into the overall avionics, and the amount of imagery to be captured on the AIR would have to be determined.

The future seems plain - cameras are the next big thing in aviation recorder technology, and the CCTV will be a growth market for an avionics industry where business at present is pretty flat. The security, safety, accident investigation, liability, and technology considerations are at a critical mass. >> Horne, e-mail mhorne@ad-aero.co.uk; Zwolinksi, e-mail mzwolinski@remotereality.com; Elliott, e-mail jim.elliott@smiths-aerospace.com; <<

Pursuing Prosecutions
Sample Incidents Where a Video Recording Would be Useful

In 1999, a former airline employee stabbed an All Nippon Airways [ALNPY] 747 pilot to death.
In 2002, near Nairobi, the fight crew subdues a passenger attempting to wrestle the controls from the pilot of a British Airways [BA] 747 moments before the aircraft was put in an unrecoverable attitude.
In a much-publicized January 2002 incident, 12 passengers on an Airtours flight from London to Montego Bay were deplaned at Norfolk, Va., following a diversion resulting from rowdy and threatening behavior. Most of the 12 individuals involved subsequently were acquitted due to lack of evidence.
In August 2002, a drunken passenger on an Air 2000 flight threatened to kill other passengers and tried to choke a flight attendant.
On Valentine's Day, 2003, a couple began a fight aboard an Alaska Airlines [ALK] flight after being "excessively romantic in public."
Source: AD Aerospace. For more incidents, see http://www.skyrage.org and http://www.thesmokinggun.com

Why RIPS? (Recorder Independent Power Supply)
Some recent accidents involving FDR/CVR power interruptions:

1998, Swissair Flight 111: No data for six minutes.
1997, SilkAir Flight 185: No data during 35,000-ft. rapid descent.
1996, TWA Flight 800: No data 40-50 seconds.
1996, ValuJet Flight 592: No data 40-50 seconds.
Source: Smiths Aerospace