How Can an Aircraft Disappear Over the Ocean? | Hacker News Reader

How Can an Aircraft Disappear Over the Ocean?

technologyreview.com

47 pointswglb38 comments

Why is there no information about the Inmarsat and Iridium networks that a lot of aircraft already use?

ADS-B is fantastic for fast, accurate position data, and apart from the hardware is essentially 'free' for an aircraft to make use of.

Satellite transceivers are also relatively cheap (a few thousand dollars), but the main expense is in sending data to and from the transceiver. This cost can be alleviated by having the transceiver programmed to send a position report from every 30 seconds to every few minutes.

It's also possible to switch between ADS-B, and satellite networks when in range of ADS-B ground stations to make things more efficient.

Satellite tracking fairly ubiquitous in Australia, New Zealand, the UK, and the US – there's even some consumer level products such as SpiderTracks [ http://www.spidertracks.com/collections/hardware/products/ha... ] which provide similar SBD modems (Iridium network) to those used in commercial aviation.

I'm guessing it's just not a requirement for aircraft operating in that airspace to use satellite tracking, as there was even less data about Malaysia Airlines flight MH 370.

Source: Worked at a company that provided both hardware and a web based application to commercial airlines to track their fleets.

That article used the word "expensive" many times, but I really have no idea what that translates to. What could be more expensive than losing a few hundred lives and an aircraft?

Does the tech used in PLBs such as the spot messenger not work over the ocean?

You have to have Boeing/Airbus approved hardware, which will come from one of the big boys in Aero Electronics (Garmin/Honeywell...) and cost megabucks due to all the redundancy, testing, engineering.

You have to have it approved by the airline, which requires all kinds of testing and red tape.

You have to take the airplane out of service in order to install it, and train tech's to maintain and diagnose it, which means updated documentation.

You have to train pilots on its functions, and how to use it properly, and all the implications of a malfunction in the hardware.

Cost != airtime + cheap PCBS slapped in a dome on top of the airframe.

"What could be more expensive than losing a few hundred lives and an aircraft?"

It seems to me like we'd still have lost these aircraft, we'd just be able to find the wreckage faster. But in the middle of an ocean that isn't currently covered, does that translate into being able to rescue anybody quickly enough to save lives? Or is it losing lives, losing an aircraft, and paying to just know where the wreckage is?

> What could be more expensive than losing a few hundred lives and an aircraft?

They aren't infinitely valuable. It's easier to understand this if you don't think of money but of labor and resources. How many days/months/years are you willing to work to provide such a feature? How much electricity, copper, oil?

(Also, satellite tracking would not have prevented the crash or loss of life).

They do. That's what they are intended for.

They even make lifejackets with then nowadays.

Pathetic.

Edit: did not see parent mentioned SPOT messenger. These have reduced coverage, as they are not emergency beacons.

Note that SPOT coverage is not worldwide, and it won't work in some parts of the world, both ocean and land:

http://www.findmespot.com/en/index.php?cid=108

It is quite good though, and would have worked for the mysterious Malaysia disappearance as well as the recent one. I'm sure other systems exist with better coverage and higher prices.

Is that same as EPIRB run on Cospas-Sarsat programme?

The coverage looks quite similar http://www.marsat.ru/coverage%20maps/GAN_coverage_I3.gif

It seems that the underlying service is run by Globalstar:

http://www.globalstar.com/en/index.php?cid=106&sidenav=232

It looks like Cospas-Sarsat is for emergency use only, while SPOT is used both for emergency location and for routine tracking.

For anyone interested in ADS-B you can dump information from nearby aircraft with a DTV stick and dump1090:

https://github.com/antirez/dump1090

I use this and it is a lot of fun. However, to do well, you need an outside antenna with good view of horizon.

Edit: And this is more fun: http://www.flightradar24.com/free-ads-b-equipment

Why not create a mesh network using the planes themselves. Surely most planes must be in line of sight with several other planes even over the ocean. That way they could broadcast their position among other things.

It sounds like ADS-B can do that:

"It does, however, have an air-to-air capability designed for use in environments like the ocean that allows aircraft to “see” each other—the FAA is now trying out this technique."

Sounds like it could work in much of the US and Europe, but planes are more sparse in other parts of the world. Africa for example has minimal ADS-B coverage and a relatively small number of planes in the air at once: https://flightaware.com/adsb/coverage

I think the author is a bit confused by how this technology works. My understanding is that an aircraft can also have an ADS-B receiver onboard, which allows the pilot to identify other traffic in the local area - it doesn't rebroadcast the signals as in a mesh network (I don't think there is any encryption in these signals either, so that's probably a good thing).

ADS-B is only really used currently as a secondary resource for identifying aircraft rather than seeing where aircraft are. The primary way is by using a radar return, and matching this up with the aircraft transponder which reports the aircraft ID and altitude.

At the moment I don't think there are any hard requirements on having ADS-B transmission equipment on aircraft, although most commercial airliners already have it (which is what FlightAware and such usually use - at least outside the US). It will become a requirement over the next few years however for most civilian aircraft.

Iridiums is rolling out new satellites that have ADS-B receivers on board (for open ocean data collection), and will be providing this data to the FAA for a fee. Its a solved problem, but will take time to roll out.

http://www.aireon.com/AboutAireon/AnIridiumInnovation

What's more, here's a product that's already on the market, and has clear pricing terms:

https://www.ybtracking.com/products-yb3

The device is ~$600, and to have a 13 hour flight transmit a GPS signal every 5 seconds would cost another $600 per-flight - I'm sure the airlines could negotiate a better deal.

Also note that they are currently available for purchase.

You can't just slap any old device into an aircraft and call it done. There's huge amounts of money + time that must be invested to get it certified, and it has to be done for each aircraft type. I haven't looked it up, but I'd wager an Iridium transmitter for transport aircraft would be in the 10s of thousands of dollars, minimum.

"instead, air navigation service providers like the FAA would pay to use" these new tools.

Seems like an appropriate government objective, but instead of just doing it themselves, govt will end up overpaying some private company with the right connections.

I don't know how may aircraft are flying over these areas of open ocean at a given time, but could some sort of mesh network be created?

Or perhaps ADS-B receivers could be placed on buoys like the ones used for tsunami detection?

It blows my mind that I can turn on my phone, go to some mapping app and get more accurate and frequent position updates than a commercial airplane. It's just absurd.

With GPS, the hard part isn't figuring out where you are, it's telling the rest of the world where you are.

A lat/long position is about 50 bits. Talk about serious bandwidth...

Oceans are huge. Where do you send those 50 bits if you are 1000s of miles from anywhere?

Satellite. The so-called cost concerns are just stupid, if I can afford a satellite phone airlines can afford to report 50 bits to satellite every 30 seconds.

With a little bit of trickery you could even easily cut that by half (e.g. drop the first few significant digits because the general area you're flying through can be assumed to be known, etc.).

Eww. So now the entire system depends on database matches back at home to work at all (since positions are relative, not absolute) and you have to test and engineer it significantly more to look for possible problems caused by hacking it this way.

Just eww. That's a terrible idea. Just spend the extra bits and save the headache of over-engineering it and then trying to fix all the problems you'll inherently create.

Yeah, I know. But the point is, it's few dozen bits every other minute - airlines, come on, Iridium can't be that expensive.

Right.

And if they don't hear you or you don't hear them, any non-zero bandwidth is a non-contact.

The phone itself is not capable of receiving the aircraft broadcasts/transmissions as they are sent at very specific radio frequency. All of what you see is actually being aggregated by a remote server from a number of receivers across the world that are capable of reading such radio frequencies, and transmitted to your device simply using Internet/3G/Network over HTTP/HTTPS.

Certainly. However, if you are here https://www.google.com/maps/place/Mariana+Trench+National+Wi... how would you communicate that to the rest of the world? That is the harder part.

Does anyone have a good breakdown for the claims I've seen elsewhere that tracking every plane would cost billions?

gps+iridium uplink ?

... and the iridium uplink, which is presumably expensive, only as a fall-back if normal HF communication is available.

Expensive relative to what? The consumer grade iridium device is only ~$1000.

Per-minute or per-byte charges are also really high. But that's "high" relative to individuals, maybe not compared to operating an airliner.

Something like 3USD/min it looks like. At GSM's 13 kbit/s, that's ~0.4cent/kbit. It doesn't sound like a significant cost to report a position regularly at that rate, not compared with the cost of flying an airliner anyway.

I'm not sure about the details of the plan, but iridium offers unlimited data for something like ~$100/month. I'm sure bandwidth is terrible, but again, these costs are small, even for individuals.

That's impressive! I severely overestimated the current cost of the service. Here's a page with good info for the curious: http://www.groundcontrol.com/Iridium_Go.htm

Reply on news.ycombinator.com