Q: As I understand it, searchers finally located the final resting place of Air France Flight 447 in 2011 and were also able to recover some of the victims and all of the flight data recorders. It seems that the freezing of the Pitot tubes were indeed the root cause of that accident and later, after the cascade of failure began, there were conflicting control inputs from the copilots until the captain realized (too late) what was happening.
Prior to working here, I used to work in the missile defense industry. We have the ability to accurately hit anything, anywhere with kill vehicles traveling at around 15,000 miles an hour. I believe these technologies can be employed on aircraft as primary or secondary telemetry data sources. I was just wondering if there has been any further discussions or developments on the subject of cockpit flight control systems advancements?
A: My review indicates that the frozen Pitot tubes played an important role in the Air France Flight 447 tragedy in 2011, and I am also convinced that the Asiana Boeing 777 crash in San Francisco (Korean Air Lines Flight 214) could have been prevented by applying the very basics of automatic safety control, which would have overruled, in one case the actions of the automatic cockpit controls and in the other the copilots’ inaction. As to Pitot tubes, in recent years, there has been some progress in converting to the use of more reliable and redundant speed detectors. On the other hand, the addition of automatic "overrule safety" controls has still not occurred, both because of ignorance and because of cost considerations.
What is meant by "overrule safety"? It refers to the automatic action that overrules all other controls, manual or automatic, and protects the system no matter what. In the processing industries, we have long applied this philosophy by, for example, providing pressure safety valves which cannot be turned off by anything or anybody. Similar "overrule safety" will probably be applied to underwater nuclear reactors, which cool automatically by thermal expansion opening and gravity-loading cooling water, without any valves or pumps. It is time for the transportation industry to also understand and accept automatic "overrule safety" controls that operate just like safety relief valves on boilers or air bags in a cars, in that they cannot be deactivated by anything or anybody.
By the way, the same applies to trains where automatic "overrule safety" controls (ATC) would also be essential. Such systems must automatically limit the maximum speed, based either just on the speed limit at the particular location or can also consider rail curvature, inertia (load on the train), push or pull mode of operation, weather conditions, wind direction, etc. The key is that it is active all the time, and its activation requires no action on the part of the engineer, nor can he overrule it.
Yes, transportation safety technology is available right now. What is missing is the willingness to make the investment needed to add the needed "overrule safety" automation. It is bordering on the ridiculous that, on the one hand, our GPS can measure the location and speed of any vehicle, or that some vendors are considering the use of automatic mini-drones to deliver pizzas, while others feel that automatically limiting the speed of trains or airplanes is too complicated or costly and can be left to bad operating controls and/or to untrained or sleepy engineers and pilots. It is the responsibility of our profession, that of the International Society of Automation, to bring this industry too into the 21st century.
A: Personally, I have found an automatic system that is on by default, but is manually overrideable when needed, to be of most value. However, I also think it really depends on the process under control, because some are just not safely (or even at all) operable in the manual mode. In any case, I think such design decisions should be made on a case-by-case basis and by persons with enough experience/knowledge of the process to reasonably evaluate the pros and cons.
Never having flown a plane myself, I would not be so sure that non-overridable auto speed is the way to go. Some type of warning of the slow speed and that the auto-speed control was only "armed" might be more reasonable.
By the way, I personally really do not like some of the latest air bag safety functions I have come across. For instance, I have been really annoyed after being stymied by the transmission position/brakes interlock when trying to restart an engine that died in traffic. And, although I realize it is not really a fault of the automatic control logic, how about those regularly failing ($900 without installation) BMW passenger seat occupancy sensors and the fact, in my opinion, that such sensors are not atypical?
A: I completely agree that we have the technology to prevent accidents like that. There is a large body of work concerning cockpit automation, under the heading of Situational Awareness. Mica Endsley has done some excellent work. Wikipedia has a good article on Situation Awareness that has lots of references to other work.