University of Bielefeld - Faculty of technology | |
---|---|
Networks and distributed Systems
Research group of Prof. Peter B. Ladkin, Ph.D. |
|
Back to Abstracts of References and Incidents | Back to Root |
This page was copied from: http://catless.ncl.ac.uk/Risks/11.78.html |
ACM Committee on Computers and Public Policy, Peter G. Neumann, moderator
_The Times_ (London), Monday June 3rd 1991. Lauda says crash engine failed New evidence that the crash of a Lauda Air Boeing 767-300 in Thailand had been caused by in-flight reversal of the thrust on one engine has stunned the aviation industry. Niki Lauda, owner of Lauda Air, made the claim in Vienna after returning from Washington, where the flight data and cockpit voice recorders of the jet were being analysed. The Austrian transport ministry supported the assertion. If the diagnosis were confirmed, the accident would be unprecedented, Herr Lauda said. The crash investigators have yet to comment. The computerised airliner's systems have the capacity for self-analysis and should have corrected such a basic error. Reverse thrust is normally locked out during flight and only used on the ground. It is thought to be virtually impossible for reverse thrust to be deployed when the engine is pushing at maximum, as the jet would have been. The incident happened about 16 minutes after takeoff. Investigators for Boeing, which manufactured the thrust reversers, were allowed access to the crash site for the first time on Friday. Herr Lauda said the flight data recorder was damaged and could not be used to analyse the crash. He said the cockpit voice recorder indicated an advisory light had come on seconds before the crash and that a voice was heard saying the light was glowing intermittently. Seconds later, First Officer Josef Thurner was heard saying: "It deployed." Herr Lauda took that to mean the reverse thrust was engaged. The entire incident, from the moment of the first warning light until the plane broke up, took no more than a minute. Last night a spokesman for the Civil Aviation Authority said that no checks were to be ordered immediately on 767s owned by British airlines.
[...] Private Speculation: Could it be something like a simple overflow/sign error to cause the engine go from max forward to max backward thrust?
[...] Now when the media make something out to be a computer error, they hardly ever are specific. What I want to know is, was it a hardware error, or a software error. And certainly, wouldn't a mid-air reversal of thrust just rip off the wing, leaving the plane to plummet down totally out of control? Any air-experts out there willing to comment on this? --Ralph Moonen
[...] However the reverse thrust system is only used on landing. It would appear unlikly (not least because of the safety aspects) that reverse thrust in NOT under the direct control of the flight management system. Also mentions mechanical interlocks, these are presumably in addition to the hydrolic systems actuating the thrust deflectors being turned off. The principle is quite simple. The thrust deflectors are locked in place unless the plane is on the ground (also landing gear is down) and the control in the cockpit is activated. Mark Evans, University of Aston, Birmingham, England
For RISKS readers interested in digital flight control systems (DFCS), I highly recommend the papers by Mackall and his colleagues on the AFTI-F16 (and some other) flight tests; [4] is particularly thorough and informative. The following extracts from [6] summarize some of the points (fairly, I hope). It seems that redundancy management became the primary source of unreliability in the AFTI-F16 DFCS. *'s indicate footnotes; they, and the references are given at the end. John - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - A plausibly simple approach to redundancy management in an N-modularly redundant system is the "asynchronous" design, in which the channels run fairly independently of each other: each computer samples sensors independently, evaluates the control laws independently, and sends its actuator commands to an averaging or selection component that chooses the value to send to the actuator concerned. The triplex-redundant DFCS of the experimental AFTI-F16 was built this way, and its flight tests reveal some of the shortcomings of the approach [2, 4]. First, because the unsynchronized individual computers may sample sensors at slightly different times, they can obtain readings that differ quite appreciably from one another. The gain in the control laws can amplify these input differences to provide even larger differences in the results submitted to the output selection algorithm. During ground qualification of the AFTI-F16, it was found that these differences sometimes resulted in a channel being declared failed when no real failure had occurred [3, p. 478]. *1 Accordingly, rather a wide spread of values must be accepted by the threshold algorithms that determine whether sensor inputs and actuator outputs are to be considered "good." For example, the output thresholds of the AFTI-F16 were set at 15% plus the rate of change of the variable concerned; also the gains in the control laws were reduced. This increases the latency for detection of faulty sensors and channels, and also allows a failing sensor to drag the value of any averaging functions quite a long way before it is excluded by the input selection threshold; at that point, the average will change with a thump [4, Figure 20] that could have adverse effects on the handling of the aircraft. The danger of wide sensor selection thresholds is dramatically illustrated by a problem discovered in the X29A. This aircraft has three sources of air data: a nose probe and two side probes. The selection algorithm used the data from the nose probe provided it was within some threshold of the data from both side probes. The threshold was large to accommodate position errors in certain flight modes. It was subsequently discovered that if the nose probe failed to zero at low speed, it would still be within the threshold of correct readings, causing the aircraft to become unstable and "depart." This error was found in simulation, but 162 flights had been at risk before it was detected [5]. An even more serious shortcoming of asynchronous systems arises when the control laws contain decision points. Here, sensor noise and sampling skew may cause independent channels to take different paths at the decision points and to produce widely divergent outputs. This occurred on Flight 44 of the AFTI-F16 flight tests [4, p. 44]. Each channel declared the others failed; the analog back-up was not selected because the simultaneous failure of two channels had not been anticipated and the aircraft was flown home on a single digital channel. Notice that all protective redundancy had been lost, and the aircraft was flown home in a mode for which it had not been designed--yet no hardware failure had occurred. Another illustration is provided by a 3-second "departure" on Flight 36 of the AFTI-F16 flight tests, during which sideslip exceeded 20deg, normal acceleration exceeded first -4g, then +7g, angle of attack went to -10deg, then +20deg, the aircraft rolled 360deg, the vertical tail exceeded design load, all control surfaces were operating at rate limits, and failure indications were received from the hydraulics and canard actuators. The problem was traced to an error in the control laws, but subsequent analysis showed that the side air data probe was blanked by the canard at the high angle of attack and sideslip achieved during the excursion; the wide input threshold passed the incorrect value through and different channels took different paths through the control laws. Analysis showed this would have caused complete failure of the DFCS and reversion to analog backup for several areas of the flight envelope [4, pp. 41-42].*2 Several other difficulties and failure indications on the AFTI-F16 were traced to the same source: asynchronous operation allowing different channels to take different paths at certain selection points. The repair was to introduce voting at some of these "software switches." In one particular case, repeated channel failure indications in flight were traced to a roll-axis "software switch." It was decided to vote the switch (which, of course, required ad hoc synchronization) and extensive simulation and testing were performed on the changes necessary to achieve this. On the next flight, the problem was found still to be there. Analysis showed that although the switch value was voted, it was the unvoted value that was used [4, p. 38]. The AFTI-F16 flight tests revealed numerous other problems of a similar nature. Summarizing, Mackall [4, pp. 40-41] writes: "The criticality and number of anomalies discovered in flight and ground tests owing to design oversights are more significant than those anomalies caused by actual hardware failures or software errors. "...qualification of such a complex system as this, to some given level of reliability, is difficult ...[because] the number of test conditions becomes so large that conventional testing methods would require a decade for completion. The fault-tolerant design can also affect overall system reliability by being made too complex and by adding characteristics which are random in nature, creating an untestable design. "As the operational requirements of avionics systems increase, complexity increases. Reducing complexity appears to be more of an art than a science and requires an experience base not yet available. If the complexity is required, a method to make system designs more understandable, more visible, is needed. "The asynchronous design of the [AFTI-F16] DFCS introduced a random, unpredictable characteristic into the system. The system became untestable in that testing for each of the possible time relationships between the computers was impossible. This random time relationship was a major contributor to the flight test anomalies. Adversely affecting testability and having only postulated benefits,*3 asynchronous operation of the DFCS demonstrated the need to avoid random, unpredictable, and uncompensated design characteristics." Footnotes 1: Also, in the flight tests of the X31 the control system "went into a reversionary mode four times in the first nine flights, usually due to disagreement between the two air data sources" [1]. 2: However, the greater the benefit provided by DFCS, the less plausible it becomes to provide adequate back-up systems employing different technologies. For example, the DFCS of an experimental version of the F16 fighter (the "Advanced Fighter Technology Integration" or AFTI-F16) provides control in flight regimes beyond the capability of the simpler analog back-up system. Extending the capability of the back-up system to the full flight envelope of the DFCS would add considerably to its complexity--and it is the very simplicity of that analog system that is its chief source of credibility as a back-up system [2]. 3: The decision to use an asynchronous design for the AFTI-F16 DFCS was because "the contractor believed synchronization would introduce a single-point failure caused by electromagnetic interference (EMI) and lightning effects" [4, p. 7] --which may well have been correct given the technology of the early 1980s. References [1] Michael A. Dornheim. X-31 flight tests to explore combat agility to 70 deg. AOA. Aviation Week and Space Technology, pages 38-41, March 11, 1991. [2] Stephen D. Ishmael, Victoria A. Regenie, and Dale A. Mackall. Design implications from AFTI/F16 flight test. NASA Technical Memorandum 86026, NASA Ames Research Center, Dryden Flight Research Facility, Edwards, CA, 1984. [3] Dale A. Mackall. AFTI/F-16 digital flight control system experience. In Gary P. Beasley, editor, NASA Aircraft Controls Research 1983, pages 469-487. NASA Conference Publication 2296, 1984. Proceedings of workshop held at NASA Langley Research Center, October 25-27, 1983. [4] Dale A. Mackall. Development and flight test experiences with a flight-crucial digital control system. NASA Technical Paper 2857, NASA Ames Research Center, Dryden Flight Research Facility, Edwards, CA, 1988. [5] Dale A. Mackall and James G. Allen. A knowledge-based system design/information tool for aircraft flight control systems. In AIAA Computers in Aerospace Conference VII, pages 110-125, Monterey, CA, October 1989. Collection of Technical Papers, Part 1. [6] John Rushby. Formal specification and verification of a fault-masking and transient-recovery model for digital flight-control systems. Technical Report SRI-CSL-91-3, Computer Science Laboratory, SRI International, Menlo Park, CA, January 1991. Also forthcoming NASA Contractor Report.
From the Boston Globe, Friday May 31, 1991 "Ask the Globe": Q. The state lottery's instant tickets now have bar codes with which sales agents can verify both their validity and the winning amount. What is to prevent agents or lottery officials from "reading" the codes on unsold tickets and taking the big prizes for themselves? A. Lottery spokesman David Ellis tells us that, once an instant ticket is "read" by a bar-code reader, it is invalidated. If a previously read bar code is reread, the computer "flags" the ticket as "previously cashed" or "cashed-stolen." The bar-code reader, Ellis adds, is "the first step in a long and complex series of protections" built into the game by Richard Finocchio, the lottery's computer manager. "The security system," he concludes with pride, "is the envy of lotteries around the world." In sending this to Risks, I am *not* suggesting that confidence in their security system is misplaced: I can think of several ways to implement this that, given acceptable security at the Lottery central computer, would prevent cheating. I can also think of several ways that would appear safe, but aren't. However, both the question and answer raise interesting questions of trust (and, as dear Pres. Reagan would say, verification). Full disclosure: I play the lottery twice a year -- when they send me their advertising freebie. In four years, I've won exactly one dollar. Martin Minow minow@ranger.enet.dec.com
> The only civilian customer for the technology has been the Australian National > Railway Commission, which at the end of 1988 adopted Viper as the core of a > new signalling system. The Australians have now gone so far with developing > the system that they would have difficulty switching to another technology. When I last heard of them (see RISKS DIGEST 6.48, 7.18, 7.36, 8.41, 9.1), the Australian National Railway Commission was suing its commercial supplier (Charter Technologies??) for precisely the same reason that Charter is now suing MoD. Peter Mellor, Centre for Software Reliability, City University, Northampton Sq.,London EC1V 0HB +44(0)71-253-4399 Ext. 4162/3/1 p.mellor@uk.ac.city (JANET)
smb@ulysses.att.com writes: > I spoke a bit loosely when I used the phrase ``probable cause''; that's a > technical term, and does not (quite) apply. Nevertheless, I stand by the > substance of what I said. The FBI is not allowed to engage in systematic > monitoring of speech without reasonable suspicion of criminal activity. We can cavil over nuances of specific terms, but smb is essentially correct. The FBI (and other gov't agencies) are generally, either by statute or policy, prohibited from routine systematic monitoring of speech outside of specific investigatory needs. This includes routine monitoring by informants or undercover operatives. Although the practice may be otherwise (see "The FBI's Misguided Probe of CISPES" by Gary M. Stern of the Center for National Security Studies), current policies are still (on paper, at least) guided by the Attorney General's Guidelines on FBI Undercover Operations" (Jan. 5, 1981) that restricts the use of undercover investigation, which is the most Constitutionally dangerous from of monitoring. There is surely nothing wrong with law enforcement agents, whether in a private or official capacity, participating in cybercommunication. The line is crossed when participation moves to surveillance, data acquisition, and record keeping as occurs in political surveillance. The Secret Service may have overstepped this line when it set up it's "sting board" (Dark Side) in 1988 in a way that seems, uh, "empirically disfactual" with the SS Director's account of it. Jim Thomas
I don't see how one can keep their anonymity in this vote by phone scheme. Passing laws against release of voter info or entrusting the security of the voter info to people makes me very nervous. Having people in the loop causes a security problem by definition. Would the people in charge of the 'voter' data center be willing to give up their right to privacy so we can monitior their activities so they can be stopped immediately if they try to divulge any voter information. Will the system be TEMPEST secured so no electronic eavesdropping is possible. Will the 'voter' data center be guarded by elite commandos sworn to defend it to the death. As I see it, if the data is stored somewhere, sooner or later the good guys or the bad guys will get hold of it some way or another either legally or illegally. Once the genie is out of the proverbial bottle, it will be too late to put him/her back. Imperfect as it is, I like the way the system is now. If people were really serious about voting, they would take the time to vote. Having a vote by phone system will only allow for more abuse and fraud as some earlier respondents have suggested. Bob Rehak, DBA At Large, BITNET: A20RFR1@NIU
This (voting by phone) is a classic example of an attempt to solve a social problem by the misguided application of technology. What, really, is the problem for which this is supposed to be solution? Invalids? Travellers? We have absentee ballots that work quite well, thank you. Low turnout? Just make election day a national holiday, as it is in most civilized countries (but not the US). Electronic voting? Who needs it? Larry Campbell The Boston Software Works, Inc., 120 Fulton Street campbell@redsox.bsw.com Boston, Massachusetts 02109 (USA)
There is a tendency to assume that computer-based voting systems such as voting by phone are more risky. I say "assume" because all election systems are risky and depend on certain procedures being carried out with integrity. But anyone familiar with the manipulation of paper or mechanical voting systems would want clarification of the statement that voting by phone relies "excessively" on the honesty of the operators of the system. In fact, allowing the individual voter to check his/her vote avoids passive reliance on the integrity of election administrators. In this connection, verification of one's vote is possible under paper voting systems and others, but computer-based systems make it easier for the citizen to initiate and carryout verification. I say this on the basis of 5 years of development work with experimental voting systems in a computer-networking environment. Grave-yard voting is a problem, but it is not clear that the problem would be worse with computer-based (phone) voting. It might be less frequent and more difficult to employ such maniupulation depending on the method of verification used. Re time-stamping: the latest schemes of time-stamping and cryptographic coding based on distributed (not centralized processing) do make it practically impossible to break the system. Put another way, the amount of information that needs to be collected to break the system is so great, that watch-guard programs would detect would-be criminals. Finally, using "none-of-the-above" is an improvement on plurality voting, which truncates information about voter preference orderings and is therefore undesirable from a theoretical point of view. This may not seem directly related to voting by phone, but Roy mentioned it as a possibility and it seems reasonable to extend the idea to consider other methods of voting that do not distort the voting process. Moreover, voting methods are related in the sense that it is not trivial to use them. I have designed interfaces for computer-based voting with different methods of voting and found what works and what doesn't work. Interfaces for voting by phone must be designed based on extensive experimentation and we may find that it is infeasible to use voting methods that have desirable social consequences. Arnie Urken
I agree with the spirit of this posting. Since we can not have complete privacy, we have at least complete access for everybody. I can readily see the day come when liberty will have as part of its definition, the freedom to access information. People who abuse the system, could have their access taken away. Much along the same lines as a robber being sent to prison. Also, a person on parole could be given varying degrees of access as their behavior warrented. PC's are becoming so cheap and powerful, that I can not see an effective monopoly being maintained for long by (m)any central power(s). The people will take back their own power. While granted, I'd prefer to have more privacy, I think it would be better to KNOW that information about me is availible as a matter of course, rather than THINK the information is private--when a privilaged subset actually has access. If we are going to have as much known about us as appears to be the case, then the only way to level the playing grounds is by letting everone have, as a matter of course, the SAME access, unless and until they abuse the trust implied by the system. If people know that their actions can be/are looked at, maybe they'll think twice about what they say or do. Granted that this is a poor sub- stitute for a proper set of values or a good conscience(sp?), but I think few would dispute that many people in power could use a constant reminder that they are answerable to the people. Brett
If you belong to TRW's Credentials Service, this is one of the "features": any time anyone accesses your credit report, they send you a nice little blurb the next day such as (from when I applied for my Gold Card): ---------------- [... assorted introductory stuff ...] The following companies have requested and received copies of your TRW credit files. Date Company Name and Address TRW Number Type of File -------- -------------------------------- ---------- ------------ 06-07-90 AMERICAN EXPRESS 3459491 Credit 4315 South 2700 West Salt Lake City, UT [... stuff about where to call with questions ...] ---------------- TRW Credentials costs $35/year. For this, you get credit card protection, unlimited copies of your credit report, and the little notes above. They also let you enter in all your financial data, so that it can be retrieved electronically with your "access number" (like a PIN) by banks and such when you apply for loans - so you don't have to fill out the forms (in theory anyway, I never tried it). Before anyone re-starts the discussion about the rip-off/non-rip-off nature of this service, it should be pointed out that: - Any time you are denied credit, you are entitiled by law to a *free* copy of your credit report from the credit bureau(s) which supplied the information to the lender. You can then dispute any incorrect info. - You can get a copy of your credit report at any time from any credit bureau for (usually) $8 and some assorted pieces of I.D. - You only get the little blurbs when someone accesses your TRW file. TRW is not the only credit bureau in the country, although it's certainly one of the largest. I believe Equifax and EDS offer similar services, though. So, unless you want the credit card protection and the little blurbs when your file gets looked at, the service is a rip-off. Me, I like the little blurbs, find the service fairly well-run, check my credit report once or twice a year, and don't mind paying the $35. --Dave Curry
[similar comment...] Isn't the market great? On the other hand, require them to do by law what they are already prepared to do for a fee? Guaranteed to cost you more. In an information society it is possible for the credit worthy to obtain the credit in minutes that was, in an earlier time, available only to the most well established in days or weeks. In part because of this, we all enjoy higher levels of commerce and a standard of living that would not have come about with out it. When we look at the effects of all of this on privacy, we should not forget the reason that we put the system in place. Yes, WE; the collective we. We did it; it was not done to us. In remedying the difficulties, we must careful not to overlook or destroy the benefits.
This page was copied from: | http://catless.ncl.ac.uk/Risks/11.78.html |
COPY! | |
COPY! |
by Michael Blume |