Aviation Security1

One main problem in the area of aviation security that might be addressed by some of the NBIC technologies would be trying to find out (a) who the people are who have access to aircraft and (b) what their intentions are.

A second problem lies in the timely detection of chemical or biological agents, particularly in airports, and in what to do about the alarms, false and real. Chemical detectors are fairly good right now, although like everything else, they can be improved, especially regarding false alarms. One does need to program them to look for the particular agents of interest. The issues then are cost, where to deploy, and how to deal with false alarms. I will touch more on biosensors in the following section.

Infotech is the technical key to determining who the people are who have access to aircraft, and it also offers the first clues to their intentions. The people with access are those who work at airports, including screeners, and the passengers and crew. One problem is to distill information from various databases, most domestic, some international, to ferret out those individuals who are known or suspected to be threats. There will be a resistance to sharing from those possessing the information on highly sensitive databases. At the minimum, a means must be found for providing only the essential information to the parties controlling access to the aircraft.

Biometrics, including facial recognition technologies, can in principle provide an additional identification tool, beyond the usual name, a minimal amount of personal data, and, perhaps, a picture. However, none of this is any use unless one has the individual of concern in one's database already. In the case of the 19 hijackers, from publicly available information, only three would have triggered any sort of alert. These were due to overstaying visas or having had minor run-ins with the law.

1 For comparison with current work, the research and development plans for aviation security within the Federal

Aviation Administration may be downloaded from the site, http://www.faa.gov/asd/red98.htm.

For those with access to aircraft, a serious background check needs to access databases that go back longer than a few months or even years: I would assert that it is necessary to track someone's credentials for eight years or more to get a clear enough picture of their potential for criminal conduct. And one constantly needs to verify that those granted access are actually the ones who have been approved for access. We don't want access given to someone who steals an ID, for example. Here, too, infotech and biometrics can only help with part (a substantial part, true) of the job. Procedural security changes are required to protect the civil aviation system adequately from the "insider" threat.

Regarding those who actually board a flight, it would be nice to know whether they have malevolent intentions that pose a risk to others. This is where some technological futurism might possibly be of use. Remote detection of heart rate, adrenaline on the skin, and perhaps other chemicals connected with the "fight or flight" reaction, is imaginable, and some efforts have been proceeding in these areas for years. Voice stress analysis is another possibility, although to my knowledge, there are no highly convincing data that this would provide a reliable trigger for the purposes considered here. And, in the neurological/cognitive realm, on an even more futuristic note, would there be clues one could obtain from a remote (at a meter or two) electroencephalogram that would be useful?

I am somewhat skeptical of all of these possibilities, but the problem is serious enough, in my view, to justify some work in these areas. At the least, one could easily imagine useful by-products for public health and neurological research. Experimental data are needed to learn how reliable (if at all) such indicators would be in a civil aviation context. The obvious issues of effectiveness, false positives, and false negatives will be determinant: a simple demonstration of some vague effect is insufficient.

One needs to bear in mind that the consequences for an individual of triggering the system may not necessary be immediate incarceration for life. A trigger may simply indicate the need to examine carefully just what the individual has brought onto the plane. One might also want to correlate alarms from different individuals on the same flight. False positives, while they need to be controlled, can be tolerated at a moderately low level (say, less than a percent).

Information technologies could obviously be applied to the issue of monitoring or controlling a hijacked plane automatically or from the ground, as has been discussed openly in the press. All this is feasible with current processing, communications, and information technologies and appears to need little further in new research. Whether this approach (especially controlling flight) is a good idea or not, is a further question. Pilots tend to think it is not.

0 0

Post a comment