amossb@umbio.MIAMI.EDU (A. Mossberg) (09/14/88)
[These are passages from the Annotation feature from Harper's Magazine, September 1988. The annotation by Alexander Cockburn (columnist for _The Nation_, _In These Times_, and _Zeta Magazine_, among others) and Ken Silverstein (freelance journalist)] The System That Brought Down Flight 655: Zeroing in on a Pentagon boondoggle, by Alexander Cockburn and Ken Silverstein [letter] - refers to a letter from John Lehman to Rep. Denny Smith [annot] - refers to Cockburn and Silverstein's annotation [letter] Dear Mr Smith, [annot] Rep. Denny Smith, Republican from Oregon, forced a new round of tests of Aegis in May 1984 after actual sea tests a year later impressed no one buut the Navy and RCA, the prime contractor. In these new tests, Aegis performed marvelously, "downing" ten of eleven drones sent its way. It was aided in this enterprise by the fact that those operating the system already knew the path and speed of the drones making up the "surprise" attack. [letter] I am responding to your..[questioning of]..the integrity of the AEGIS system... [annot] In the rush to judgement after the U.S. downing of..[flight 655] ..no guilty party was more carefully hidden in plain sight than Aegis...This most costly package of electronic complexity mistook an Airbus (length: 175 feet, 11 inches) for an F-14 (length: 62 feet, 8 inches), miscalculated the altitude of the plane by some 3,000 feet, and determined that the Airbus was descending when it was actually climbing.. [letter] Reports of inadequate testing of TICONDEROGA and its AEGIS system are not correct. AEGIS is the most carefully tested combat system ever built.. [annot] The "care" was mostly exerted to deceive Aegis's many critics. In one important series of tests, the Navy set up components of Aegis in a meadow near Exit 4 of the New Jersey Trunpike--"operational tests," the Navy called them. There Aegis performed such difficult tasks as monitoring the comings and goings of civilian air traffic over New York airports. As the Navy well knows, radar reflections off land are entirely different from those off water, and meadows don't pitch and roll--but then the object of the exercise was not any serious testing of Aegis but the extraction of more funds from Congress. [letter] TICONDEROGA establishes an entirely new level of sea-borne anti-air warfare capability -- it is the first Navy ship desgined to counter large scale anti-ship missile (ASM) attacks. [annot] To counter missile attacks, ships such as the Ticonderoga and the Vincennes utilize the SLQ-32, an electronic system allied to Aegis that works like a Fuzzbuster--it identifies approaching ships and planes by the radar they emit. The SLQ-32 on the Vincennes, however, was unable to distinguish between the weather radar aboard an Airbus and the combat radar installed in an F-14. [letter] The fast reaction time and high fire power of the AEGIS weapon system have proven successful in countering ASM threats in operational testing against the lowest flying and fastest targets available to the Fleet. [annot] Nonsense. Aegis is particularly inept at detecting planes and missiles at low altitude, the most likely path of any agressor. In two 1983 tests, Aegis missed six of seven low-altitude targets. [letter] TICONDEROGA, acting in concert with other battle forces, provides a quantum leap forward in the defense in depth capability of our carrier battle groups. [annot] Aegis in fact imperils every ship on which it is installed. The system emits four megawatts of energy--equivalent to 40,000 100-watt light bulbs--the moment it is activated, turning the ship into a powerful electronic beacon and making it an easy target, especially for the radar-homing missiles the Soviets have sensibly developed. [letter] I appreciate your support in a strong Navy. [annot] Support indeed--each Aegis costs the taxpapers about $500 million, half the cost of the cruiser on which it is installed. [letter] Sincerely, John Lehman [annot] John Lehman--who as Secretary of the Navy was the patron and close friend of Melvyn Paisley, the key figure in the current procurement scandal under FBI investigation--thought the expense of the Aegis system "entirely justified." Admiral Thomas Davies, who before retiring tried to kill the program, calls the system "the greatest expenditure to get the least result in history"--which should be adopted as the motto of today's military procurement system. A pair of binoculars could have told the officers of the Vincennes what was flying overhead. But binoculars don't cost half a million dollars. The more complex the weaponry, the deeper the pork barrel and the more swollen the bottom line. This is the system that produced Aegis, and did in the 290 passengers aboard Flight 655. aem
dtynan@sultra.uucp (Dermot Tynan) (09/17/88)
I don't really want to drag up this whole question again, but it seems to me we're missing one important point. Accountability. It may very well be, that the AEGIS system is a pile of junk. To me, that's not the issue. The fact that someone would shoot down an aircraft based on a computer analysis is scary. It reminds me of "1984", and future-shock-type SF. As an example, I can remember my ex-bank (and I mean EX), telling me I was overdrawn when I knew I wasn't. I spent fifteen minutes trying to persuade the teller that there was most definitely money in my account; "But Mr. Tynan, the computer says you're overdrawn"; "Well, the computer is wrong!" (for this comment I got a disbelieving stare :-). Eventually, what won the issue for me, was telling her "MY computer says otherwise. Why don't I have it call you?". (As it happened, my computer at the time was a small CP/M system that knew less about my bank account than my calculator). When she went back and checked the paper receipts she found a couple of deposits uncredited (!!). What amazed me was her comment; "Well, the computer is about two weeks behind, because of overwork". Anyway, I think there is a definite trend on our side of the industry (by OUR, I'm not referring to defense companies, but computer people in general) to add "features". I can imagine two software people late at night, working on AEGIS, thinking... "Lets add a feature to try and analyze the data, so that we can tell the operator what kind of plane it is". Such a system is interpretive at best. And NO WAY can it be foolproof. On the other hand, the actual effect on the part of the operator, is to assume that the *computer* knows best. Hah! I once got a letter from a well-known credit-card company because of a past-due balance. The letter was signed by Mr. Reeves (I can't remember the actual name). I can't prove this, but when I asked to speak to him, I was told he "didn't exist", and that this was a message put out by the computer. Boy, is that scary! As a group, we should have a philosophy of education for people not directly associated with the technology. The two major concerns I have, are; a) bringing back accountability, and b) teaching people to trust computers as much as they'd trust the people who enter the data. I think that these failings will account for a slow-down in the way people use computers, and in the long-run will affect our positions and credibility. We've all heard the jokes about the computer sending out the bill for $0.00 and so on. These jokes reflect an underlying fear of the technology. This has got to change. I apologize for the length of this comment, but not it's content. Regards... - Der