Robot Submarines Seek A Downed Plane's Secrets

A trio of deep-sea robots are the best hope yet for finding what caused one of aviation's most mysterious disasters.

Air France flight 447 left Rio de Janeiro at 7.03pm local time on 31 May 2009, heading for Paris. Three hours later, over the mid-Atlantic and beyond radar range, it entered a series of severe tropical storms—and never emerged.

In March, a team of sub-sea search specialists will be taking three autonomous submarines to search the sea floor beneath the area where parts of the plane were found. What they most want to find are the flight recorders, to help piece together what happened that night. Previous attempts have proved fruitless, but the submarines now have cutting-edge upgrades that the team believes give them a far better chance of finding the wreckage.

The expedition, which will start on 18 March, is charged with emotional significance, a fact not lost on team leader David Gallo of the Woods Hole Oceanographic Institution in Massachusetts (WHOI). "We recently met some of the families and loved ones of the victims," he says. "For their sake, not just the flying public and the aviation industry, we're now extremely motivated to find out exactly what happened to this aircraft."

All that is known of the Airbus A330's demise comes from its last communications: a series of automated messages from the aircraft's computers, chirped via satellite to an Air France maintenance centre just over 3 hours after take-off. They warned that inconsistent airspeed data was being fed to the flight computers.

Search teams turned up little else, finding the bodies of just 50 of the 228 people on board and small amounts of floating wreckage—including the plane's severed tail fin—1000 kilometres north-east of Brazil's Fernando de Noronha islands. To date only 3 per cent of the aircraft, by weight, has been recovered.

Crucially, the two black box recorders remain missing. The team will be searching for the shoebox-sized devices in a 10,000-square-kilometre area stretching across the mid-Atlantic ridge, a vast undersea mountain range that has stymied three previous expeditions. Volcanic boulder-covered ridges, some with near-vertical slopes, rise to within 600 metres of the surface. Troughs can plunge 4000 metres into the abyss.

One way to find objects on the sea floor is to use a ship towing a sonar array. This provides an acoustic image of the seabed. The problem with the mid-Atlantic ridge is that the sonar has to "fly" so high to avoid its tether snagging on ridges and boulders that the resolution is very poor.

Last year, in an attempt to get around this difficulty, WHOI used its Remus 6000, an autonomous underwater vehicle (AUV) that has its own, high-resolution sonar scanner. It searched for the flight recorders on pre-programmed routes, but kept having to stop scanning while it manoeuvred around the many ridges.

The sub has now been upgraded, says WHOI senior engineer Mike Purcell. Its sonar system uses sound waves with double the wavelength of the original, meaning it can generate an acoustic image from 130 metres above the sea floor instead of 90 metres, at the same high resolution. "In rugged terrain with sharp changes in depth we will maintain bottom tracking more of the time," Purcell says. This time the WHOI is sending three of the torpedo-shaped Remus 6000s.

The subs' autonomous guidance software has also been rewritten to increase their dive angle from 15 degrees to up to 25 degrees, so they can scan more of the steep ground that was previously missed.

Improved control software will reduce the chance of sea-floor collisions, and object recognition software will enable the subs to automatically identify unusually shaped objects as "interesting" and drop to within 7 metres for a closer look with their cameras.

The team will use the upgraded AUVs in staggered shifts, running them 24 hours a day from the expedition ship MV Alucia in three 30-day search programmes.

Finding the flight recorders would help families seeking compensation, says Peter Neenan, a lawyer with Stewarts Law in London, which is prosecuting a case for the families of 50 AF447 victims in a court in California. The claimants allege that equipment supplied by several firms, including Airbus and Thales, was "defective and unreasonably dangerous in one or more respects". Thales makes pitot tubes—the instruments that intercept airflow to measure airspeed.

Interim findings by the BEA, the French bureau for civil aviation security, said that icing of the aircraft's three pitot tubes could have led to confused speed readings. These might have crashed the flight computers or misled pilots flying manually, risking the critical loss of lift known as a stall.

Still, it is not yet clear who is at fault, if anyone. "Finding the recorders would make things a lot easier. We've got our fingers crossed," says Neenan.

Airbus, which along with Air France is funding the search, also wants the recorders: "Unreliable speed data from pitots alone would not have led to what happened—but it could have contributed," says spokesman Justin Dubon.

If the recorders are located, the French government has agreed to finance their retrieval. It might also fund recovery of the remaining bodies, depending on the state and location of the wreckage.

Neenan says the AF447 families desperately need answers and closure. Ironically, another Atlantic disaster may help: WHOI has just finished mapping the debris field around the wreck of the Titanic with a Remus 6000. The sub spotted items as small as pots and pans from 600 metres away, lending hope that AF447's elusive recorders might yet be found.

New Scientist reports, explores and interprets the results of human endeavour set in the context of society and culture, providing comprehensive coverage of science and technology news.

Image Credit: Wikipedia


Comments

    Thanks for a good concise article.

    The recorders will be recovered once the major debris field has been identified. Large parts of the hull, wings and engines etc.. are the items first expected to be located. This will be no small feat in itself, as to find what remains of the aircraft is like looking for the proverbial needle in the haystack, or in this case some 17,000 square kilometers - even worse by comparison.

    Best of luck to the WHOI's REMUS team and the crew of M.V."Alucia".

    >risking the critical loss of lift known as a stall.
    Student pilots practice stalls by their third lesson. Even a high speed stall is recoverable.
    What would have been bad for AF447 is a deep stall wherein the control surfaces are already fully stalled and therefore control inputs by the pilots are useless. This would be most likely to occur in an overspeed situation wherein the plane's Fly-by-Wire system disengages and attempts to impose Alternate Law (wherein the pilots fly the plane without computer override), but such Alternate Law mode fails due to massive electrical buss overloads.
    Whether it was a flat spin or not is unknown but its likely that there was very little forward motion of the airliner, most of the motion would have been vertical.
    The problem is that when the computers suddenly turn themselves off and the pilots are required to fly the airplane, those pilots often have very little experience actually flying an airplane rather than inputting commands into a computer that does the flying. There is no "learning curve" during an incident wherein most instruments are being lost and whatever instruments remain are unreliable. Any student pilot knows to Lower The Nose and Add Power, but during conditions of high informational overloads on the airliner pilot its difficult to get AutoPilot and AutoThrust out of their minds and actually start flying the airplane themselves.
    In Fly By Wire the computer overrides the pilot's unsafe actions, in Alternate Law mode it does not. The problem is that few pilots have any actual experience in Alternate Law mode.

    Is it possible to Fly these planes with the computer turned off? IE take-off, fly, land?

    Instead of the Thales open pipes freezing up and speed monitoring lost.
    Could not pressure sensors be fitted to the windscreen that detect and measure wind pressure like a condensor microphone works? (Variable capacitor).
    I assume the windscreen does not ice up and is kept heated to prevent misting up.

Join the discussion!

Trending Stories Right Now