What I had understood before was that the airplane carried an Inmarsat satellite transceiver which remained active, even though certain subsystems that use it (notably ACARS) were disabled. While active, the transceiver communicates with the satellite once an hour (the "pings" at 11 minutes past), even if it has no traffic to send. Failing to disable the transceiver in addition to the other systems would have been an oversight on the part of the plane's hijackers, assuming it was hijacked.
Apparently, Inmarsat had available accurate ping timing data. Combined with knowledge of turnaround delays, this would enable them to calculate the plane's distance from the satellite during each ping. That defines an imaginary sphere in space, which intersects with the earth's surface to form a circular locus of where the plane could have been at the time. Presumably, there were eight such ping interactions, ending with the one at 0811 Kuala Lumpur time. There was no 0911 ping, meaning the plane had either crashed or landed and powered off before that time.
And so we had the northern and southern arcs. These were obtained by cropping the circle based on factors such as how far the plane could have flown in the available time, whether or not military radars, if any, had detected the plane, etc.
Now comes the Doppler data. Apparently, Inmarsat also has an accurate read of the plane's transmit frequency. The extent to which this frequency differs from the nominal frequency would indicate the plane's velocity towards or away from the satellite at the time of the ping.
Today, Chris McLaughlin revealed that Inmarsat had studied the Doppler patterns of normal flights in the same region and concluded that the Doppler data for MH370 indicated that the southern arc was the right one. Hopefully, Inmarsat has it right, and the black boxes will be found before their beacon batteries run out.
One minor correction, which I'd seen an indication of and had suspected from an architectural design point of view: the plane was getting pinged, it wasn't producing the pings. Per The Aviation Herald, which has lots of details today (and isn't behind a pay wall who's visits per month I used up yesterday): http://avherald.com/h?article=4710c69b&opt=0
"During the flight the ground station logged the transmitted and received pulse frequencies at each handshake."
That's in the middle of a huge paragraph describing how they did it in quite some detail, e.g. compared data to other 777s flying that day.
No. Persistence of the data does not require continuous power. But the black box has a battery-powered "pinger" that produces sound to make it easier to find. The batteries for that is only required to last for 30 days.
For communication under(salt)water, you're pretty much screwed with every part of the electromagnetic spectrum apart from visible light, and even that won't work in cloudy conditions or if the emitter is occluded. Sound is really the best option.
Because infrared doesn't work very well in water, and crashes in water are the biggest challenge for finding wreckage.
Airliners also carry emergency locator transmitters† (ELTs), which, when activated by a crash, transmit radio signals which can be triangulated via satellite. However, ELTs do not work underwater. No ELT signals from MH370 were heard.
> It seems very clear that these airplane blackboxes are ridiculously under-specced and underpowered.
Sorry but you really don't know what you are talking about. Black boxes are not made to be latest generation hardware and top class storage equipments, they are made to resist and survive and air crash as their first priority. And in most cases (i.e. crash on land, since most crashes occur at take-off or landing) 30 days is far enough to recover the data and the black box.
Note that even without the beacon, the AF Rio flight black boxes were recovered 2 years after the crash, under extreme depths of water.
> Black boxes are not made to be latest generation hardware and top class storage equipments, they are made to resist and survive and air crash as their first priority.
You could say the same about, say, the body of the airplane or the engines. But we have innovation in those fields (composite materials, exotic metals, etc.).
The point is: 2 hours is simply not enough. The BB should be able to hold the entire flight's data; and our current technology (in terms of processing power and storage capacity) is more than enough to do that. Plus, given that the BB is a modular, independent unit it would be easy to swap out an older BB with a newer, better BB.
Why 2 hours isn't enough? We're talking about voice conversation between the crew. Mot of the information meaningful to investigation would happen in a short period of time prior to the crash.
Usually correct I imagine but what about the slow de-pressurisation scenarios where the pilots have passed out and the plane flies until it runs out of fuel?
In this case (whatever the cause) it would definitely be useful to have the recording for the time the plane went off the planned route but that will probably have been overwritten.
That's a good point. However I believe low pressurization event is captured by the data recorder. That information, coupled with other events, and the fact that the voice recorder goes silent for 2 hours, would give investigators sufficient clue.
I don't have any inside knowledge of how blackbox works the way it does, but I imagine, like with any critical system, reliability and redundancy trump everything else. I bet the two hour limit is not an arbitrary number but rather result of some lengthy tried and true testing process and debate. Any design decision is a trade-off, no matter how small.
I'd argue that, if we have learned anything from MH370, ensuring blackbox survive in the harshest environments in the planet and stay discoverable for long period of time should be the utmost concerns.
Why can't the damn thing record audio (and lots of other data) from the entire flight, rather than only two hours?
I commented on this on a prior thread [1]. There has been push back from the Airline Pilots Association against longer recording duration on privacy grounds:
The Air Line Pilots Association (ALPA) did not support the proposal to increase CVR recording time because the FAA did not propose any increase in the privacy protections regarding the access and use of information recorded on a CVR. The ALPA stated that existing protections are inadequate despite years of its attempts to change the standard.
>>Why not have an atomic (nuclear) battery that will be able to send out sonic pings for a year?
Because atomic batteries use RTG cells which convert heat directly into electricity - and that's one of the most inefficient ways of doing it, which means, that for amounts which would be safe to put in an airplane, you would have maybe a couple watts of power, and I doubt you would even get that much. Also, nuclear batteries don't last forever, for them to be producing power a highly radioactive materials must be used,with super-short half lives - a material with a half life of hundreds of years doesn't produce enough heat through radioactive decay.
Which means that those batteries would have to be frequently replaced - and as with anything nuclear, the costs would be enormous. 30 days to find the black box after a crash is plenty.
Atomic batteries are not at all powerful enough. Microwatts.
Running anything on batteries for days takes a lot of batteries. A Raspberry Pi for 2 days for example takes somewhere from 1.5 to 2 kg of li-ion batteries.
That sort of weight starts to be really problematic for a device which needs to survive a plane crash - since every bit of extra mass increases the force with which it hits.
They don't have to be microwatts. Galileo, for example, used 17 pounds of plutonium to generate 570W of power at launch, with a half life of about 88 years.
Not that you'd want to carry 17 pounds of plutonium around on an airliner all the time, but power output is not an insurmountable problem by itself.
Sticking a bunch of nuclear material on airliners to aid in accident investigation seems kind of counterproductive. Imagine how much more fun the 9/11 cleanup at the Pentagon and WTC would have been if the workers got to play "find the needle in the haystack" with a couple of pounds of plutonium along with all the rest.
I don't understand why the pingers are on continuously until battery depletion. Assuming we can't have them be triggered by incoming sound (making them essentially sonar transponders), why couldn't they be on for e.g. 15 days, then off for a month, on for 7 days, off for two more months and on for 7 days? The power requirements of the timing device are minuscule (my watch needs a new battery every few years).
How would turning the pinger off for a month help? It just means that the search operation stops for a whole month while they wait for the pings to resume. I don't understand how your proposed idea would benefit a rescue operation.
I'd suggest something more sensible. E.g. one ping every Ns for the first 15 days, every 2Ns for the next 15 days, every 4Ns for the next 15 days... oh look, "infinite" duration.
However, I suspect the "pinger" mechanism is very, very simple and completely analog. No logic in its circuit at all. Because having it SURVIVE to ping 30d was the key consideration.
the second biggest issue with the data is that it will have written over some of the information that would have occurred at the time the flight deviated from the assigned course.
It used to be that voice recorders with 30 minute loops, I am not sure if that is still the case.
I suspect the answer is similar to "why don't they stream FDR-data, live, for all flights?".
That being: the airlines feel the cases in which extra measures matter are only a small subset (flights with problems in incredibly remote areas) of a small subset (flights that have any problems at all) and thus not worth the extra cost/complexity/risk to all the other flights.
engineers were able to narrow the location of the plane down
with an initial analysis of the Doppler effect on the signal
from the flight’s pings and the plane’s approximate altitude.
How and why would they have a record of this to be able to study it?
I'm curious about that too, it seems very unlikely that the actual digitized RF waveform was available for analysis. My guess, based on working with somewhat similar systems in the past, is that the ping packet content was also tagged by the satellite with the measured carrier frequency (or offset from nominal) and this value had enough precision that it let them calculate the Doppler effect. The reason I would say this is because I've done similar things in the past for telemetry applications, it can be really handy for debugging when things aren't working right.
Inmarsat is a "bent pipe" system, meaning that the satellite just amplifies and sends down to the base station everything that it receives. So the operator's RF fanciness is limited only by the noise factor in the satellite's amplifiers.
Ah, that makes a lot more sense, though nevertheless impressive precision on the part of the base station.
However, given the corridors, it still seems like the suggestion is that the Doppler measurements said it was moving rightward or leftward, when it should only say how much the aircraft was getting closer or getting further away. Looking at the map alone, it doesn't seem like that information would help distinguish which corridor it was in.
The satellite was also moving (south or north), and the measured redshift/blueshift could be used to determine if the plane was north or south of the satellite.
Geostationary satellites don't actually stay completely stationary for any length of time without active stationkeeping. They very quickly develop a north-south motion which makes them trace out an analemma, a sort of figure eight on the ground.
I first wrote geosynchronous and approximately geostationary, then decided it was too cumbersome. What sort of magnitude does the analemma have? Also is the effect amplified by it being a "bent pipe"?
Even if it isn't intentional, in scenarios such as measuring a Doppler shift I can see how it can be a useful tool. Almost analogous to microsaccades we make with our eyes.
What is a likely way for that to work with something launched in 1996? If the plane was going 50 MPH away or toward satellite, that's ~7e-8 * c. If I remember, the frequency would be shifted by a factor of the square root of (1 - 7e-8)/(1 + 7e-8), or its inverse.
Would the satellite always relaying that level of precision to the ground? If it is, why would they have prioritized that over what would have presumably be a dramatic increase in capability if it were discarded instead? Could they have designed it to store that much precision in rotating logs available on request? And furthermore, how is a relativistic Doppler shift on those magnitudes discernible from the frequency shift naturally occurring from the plane's transmitter?
I suppose my point is that while it's easy to understand the idea of a train whistle changing pitch as it passes by, how they would do it in practice sounds like magic, being unfamiliar with the capabilities and the equipment that geostationary satellites generally carry.
Apparently it was hourly and Inmarsat calculated two arcs (north and south) which it could have taken, but then focus seems to have moved to the southern one in last few days.
I wonder if they flew test flights along both arcs and compared the data.
They didn't need to fly test flights. That would be big $$ and take time.
But apparently, their hardware is installed on plenty of airplanes, so they just needed to study normal flights north and south in the region to establish a baseline to which to compare MH370's data.
And it was that data that lead them to conclude that the southern route was likely the one MH360 actually took.
I should have been more clear -- they compared routes and aircraft velocities that could be expected to produce the same radial velocity and Doppler effects, given the satellite's geostationary position.
So the idea is that there's a shift on the intra-ping timescale? So there are numerous packets and each packet would have the frequency of that particular packet attached to it? Obviously I have no clue what the protocol for the ping looks like, just trying to understand how it would work.
Also curious how they matched it to planes flying the southern route, since I wouldn't have guessed there were any other planes flying on the southern route (which goes...nowhere).
Logged ping, base carrier frequency 1375.003762 MHz.
Logged ping, base carrier frequency 1375.003761 MHz.
Logged ping, base carrier frequency 1375.002961 MHz.
Then with some maths and comparing to the known positions of the plane (pre-contact loss) and known positions of the satellite you can work out which way the plane would have had to move to get the observed results once contact was lost.
(numbers in example made up, and likely aren't appropriate for satellite communications but you get the idea)
Well sure, but I don't know if doppler information between pings buys you much new info.
The pings, as I understand it, came an hour apart, and continued for 7-8 hours total. With just the pings themselves they were able to extract the distance from the sat, which is what led to those arcs they published last week. I dunno if doppler shifts from pings an hour apart add much to that.
Conservatively there are tens of thousands of commercial flights per day. They could use each flight as a test case to understand and refine the model of the original analysis and come up with something more accurate.
In the actual video he seems to have misunderstood what the Doppler effect actually is. He gives the example of a train whistle "getting louder as it comes towards you and fainter as it moves away" but he should have said something along the lines of "the same way that a train whistle has a higher pitch as it moves towards you and a lower pitch as it moves away from you". The intensity of the signal is much harder to measure/calibrate than the frequency of the signal.
...and by the way, we offer a service that only costs $1/h/plane and we'd like governments to force everyone to pay us. Now, I have no idea whether $1/h/plane is reasonable (is that to cover the satellite comms, or is that on top of them?) but it is kind of stunning that planes aren't tracked automatically and as a matter of course the way, say, most people with cell phones are.
Call me a conspiracy nut, but here's what I believe may have happened:
It's almost a certainty that the US government has spy satellites video recording the entire planet at all times, at a level good enough to distinguish individual aircraft. The US spy agencies probably knew where the plane was right away, but they can't just tell everybody or the jig is up.
So they may have just contacted this satellite company, told them where the plane is, and came up with a technical explanation of how it was located.
The US certainly has a bloated defense budget, but I strongly doubt that we have satellites constantly trained on the typically empty oceans that cover 70% of the Earth.
While I don't subscribe to GP's view that satellites have bandwidth for "video recording the entire planet at all times", it seems plausible that SIGINT satellites would have a relatively constant bandwidth available for observation. As a (non-geostationary) satellite passes over relatively unpopulated areas (e.g., the southern Indian Ocean), it will be "bored" and thus far more likely to notice a single low-power transmission.
On the other hand, if it is storage and downlink bandwidth that is limited, sats would probably be programmed to ignore or discard uninteresting signals. This seems likely to me.
So the question may be whether commercial airliner pings from the southern Indian Ocean area were considered 'interesting' at the time and, if not, did anyone realize there was a rogue 777 in time to re-task the satellites?
Regardless, I'd bet those signals are considered interesting going forward.
The other complication is that Inmarsat up-link is a steered spot-beam ( the antenna on the aircraft is stabilsed and driven, or phased-array in recent implemenations ). So the SIGINT sat would have to be passing through the up-link 'beam' precisely as it was transmitting.
Down-links are ( depending on generation and usage ) either from one of an array of narrow-beam antennae on the Inmarsat or from the wide-area general broadast antenna; much easier to intercept, you and I can do it with a sat-TV dish and an RTL dongle.
Brochures I saw years ago indicated that least some portable Inmarsat terminals are also directed beams. Taliban/AQ was known to use directional antennas for communications security, at least later in the war. So possibly we could say the same thing about Bin Laden's satphone, and we know how that turned out.
Call me a conspiracy nut, but here's what I believe may have happened:
It's almost a certainty that the US government has spy satellites video recording the entire planet at all times, at a level good enough to distinguish individual aircraft. ...
You are a conspiracy nut. :-)
Spy satellite time is expensive. And they typically have a narrow field of view (the better to read the license plate number off of your car). Basically they are orbiting telescopes like the Hubble, except they are pointed down instead of up.
It is inconceivable to me that they'd waste time having a spy sat looking at empty stretches of the Indian Ocean for no good reason. Especially when there are more interesting things to look at (Afghanistan, Ukraine, North Korea, etc.).
While I don't think this particular suggestion is likely given the shear amount of data that would need transmitted, it might be possible that there is more data being transmitted by these planes than has been admitted.
based on the region where this occurred, you can rest assured that one, if not several, state intelligence services knows exactly what happened to the plane and has near continuous knowledge of its position. the area of asia where this occurred would likely be of interest to US, CN and RU.
the fact that none of these countries gave information about the whereabouts of the plane in a timely fashion suggests to me that they have a vested interest in not sharing what they know.
What I had understood before was that the airplane carried an Inmarsat satellite transceiver which remained active, even though certain subsystems that use it (notably ACARS) were disabled. While active, the transceiver communicates with the satellite once an hour (the "pings" at 11 minutes past), even if it has no traffic to send. Failing to disable the transceiver in addition to the other systems would have been an oversight on the part of the plane's hijackers, assuming it was hijacked.
Apparently, Inmarsat had available accurate ping timing data. Combined with knowledge of turnaround delays, this would enable them to calculate the plane's distance from the satellite during each ping. That defines an imaginary sphere in space, which intersects with the earth's surface to form a circular locus of where the plane could have been at the time. Presumably, there were eight such ping interactions, ending with the one at 0811 Kuala Lumpur time. There was no 0911 ping, meaning the plane had either crashed or landed and powered off before that time.
And so we had the northern and southern arcs. These were obtained by cropping the circle based on factors such as how far the plane could have flown in the available time, whether or not military radars, if any, had detected the plane, etc.
Now comes the Doppler data. Apparently, Inmarsat also has an accurate read of the plane's transmit frequency. The extent to which this frequency differs from the nominal frequency would indicate the plane's velocity towards or away from the satellite at the time of the ping.
Today, Chris McLaughlin revealed that Inmarsat had studied the Doppler patterns of normal flights in the same region and concluded that the Doppler data for MH370 indicated that the southern arc was the right one. Hopefully, Inmarsat has it right, and the black boxes will be found before their beacon batteries run out.