The Internet of Flying Things gets some extra attention in the wake of 2014’s aviation disasters
via Flickr © dougkeeney (CC BY 2.0)
It hasn’t been a very good year for aviation with the unprecedented loss of three large passenger jets - the only obvious common denominator was that they all belonged to Asian carriers. Malaysian Airlines MH370 simply and literally disappeared off the radar early last year and still hasn’t been found; Malaysian Airlines MH17 was shot down while flying over the Ukraine; while the most recent, Air Asia flight ZQ8501, ‘appears’ to have hit catastrophically bad weather (the black boxes have yet to be found, though the remnants of the plane have).
All three disasters appear to have different causes and are being seen as an unfortunate statistical cluster rather than evidence of some specific malaise that might be fixed. But while we might not have a common cause, what the incidents have highlighted is the inherent limitations of the ‘black box’ technology which has traditionally provided the evidence for crash investigation: a marvel of 20th century technology that’s now well past its sell by.
The black boxes just don’t cut it any more. First, they sometimes can’t be found (as the first crash showed), but most importantly, while they might be able to tell how a disaster happened after the fact, they don’t do anything clever and 21st century-like by being interactive, or at the very least by providing real-time communications back to air-traffic control. If they did it’s possible that the first incident (when MH370 was presumed to have crashed into the Indian Ocean ) could have been averted completely, while the plane involved in the most recent accident - ZQ8501 - would have been found much faster if its co-ordinates, altitude, attitude and other data were being constantly monitored.
IATA, the international regulatory organisation for commercial air travel, set up an Aircraft Tracking Task Force (ATTF) in the wake of the loss of MH370 and it’s just reported back and action is likely to be taken to develop technology and protocols to track aircraft in real time - continuous tracking. At present aircraft travelling internationally tend to be tracked by individual traffic controllers when over national territory but then spend much of the flight off-radar (as it were). Continuous tracking is likely to be implemented.
But other voices are calling for a much more. Continuous tracking might have prevented MH370; nothing could have averted MH17 which was brought down by a missile; but there remains the weather problem encountered by ZQ8501.
We’ve all been in planes when sudden and unexpected turbulence intrudes, usually just as you’re embarking on a meal. Sometimes it results in a bumpy ride only; sometimes the pilot will take evasive action by dropping a few thousand feet to get away; and sometimes - very occasionally - the weather will hit so hard it causes the sort of catastrophic failure presumed to have happened to ZQ8501.
Consensus is that it’s not good enough to issue general weather guidance to pilots and then expect them to take evasive action AFTER they hit nasty weather. Much better to develop a more granular understanding of weather conditions so that pilots can actually steer around trouble-spots. The key is to use the planes themselves to gather the data.
Outfits such as Microsoft Research are working on how to collect and analyse wind-speed data already collected by US planes in near real-time. By pulling together this data and mashing it with existing sources of US weather data the researchers claim they can get a bead on very local weather conditions with greater accuracy than supercomputer models can predict ahead of time.
This is IoT and ‘big data’ working together to perform a valuable real-world application.