Three days of safety talks -- Wrap up from Japan

My sister toasts Capt. Kohli's excellent
presentation at the ISASI conference

With all the 21st century tools and technology, very few air accidents go unsolved anymore.  But  that is not to say that determining with the cause of an accident is easy. On the contrary sometimes the easy answer is the wrong answer and no one knows that better than Samir Kohli, head of safety for the Saudi Aviation Flight Academy. 

When a pilot with the Indian Navy crashed a helicopter on the deck of a carrier during military exercises a few years ago, the immediate assumption was that mishandling of the controls caused the tail boom to strike an obstacle sending the aircraft into a spin. Only when the pilot asked Capt. Kohli to represent him during a military court-martial did a closer examination of the wreckage, and the injuries of those on board the helicopter reveal the failure of a rotor separation cable led to the accident. 

But more than the exoneration of one pilot was at stake.  An inspection of helicopters in the fleet  showed similar  cracks of the same part on a number of helicopters.  Had Capt. Kohli not challenged the “easy” investigation, who can say how many more accidents might have occurred?

Capt Kohli’s compelling story was one of a number presented at the International Society of Air Safety Investigators' annual conference in Sapporo, Japan that hewed to the theme that for investigations to be worth the investment, they must be done right.  And it can be hard to resist the allure of the quick and simple resolution. Especially in the absence casualties. 

The dramatic landing accident of a British Airways  Boeing 777 on a flight from China to London in January 2008 was, miraculously, fatality-free. For the past two years, investigators from several governments and aircraft and engine manufacturers have been energetic in their efforts to to learn more about engine ice, suspected to have caused both engines  on the widebody to rollback so that they were not producing enough thrust during the critical landing phase resulting in an accident.

Luke Schiada, Lorenda Ward and 
John Lovell of the NTSB -USA

The puzzling nature of the event, and the fact that it was followed by a similar episode on a Delta Airlines flight later that same year prompted the National Transportation Safety Board and the Federal Aviation Administration to issue emergency recommendations and directives.

"I could spend the rest of the day talking to you about ice," said Brian McDermid, of the UK office of the Air Accidents Investigation Bureau who was assigned to find out what happened to the British Airways flight. "We know relatively little about ice in fuel." 

McDermid, Tom Dodt, and Mark Smith engineers from Boeing presented papers on their work describing how they created a full scale 777 fuel system to study how threatening ice could form during flight.

"Every long haul flight experiences ice," Smith said he concluded. And he reminded the 200 investigators in attendance that most of them would be taking a long-haul flight back to their home countries in the next few days. "Its an industry issue."

The search for understanding of fuel system icing looks forward but also back in time. McDermid's research turned up documents dating back to the 1940s. That fact was uncomfortably familiar to me. While writing my book, Deadly Departure, about the 1996 crash of TWA Flight 800 from a fuel tank explosion at 13,000 feet, I also found a problem - fuel tank flammability - that had been studied for decades without significant progress towards resolving the threat.

Hiroto Kikuchi (right) with USC 
Viterbi School of Engineering's Michael Barr

One of the more intriguing presentations at the conference was by Hiroto Kikuchi of the Japan Air Self Defense Force, a psychologist and investigator who analyzes voice patterns and tone to determine pilot stress levels. Kikuchi suggests this research can be used to glean additional information from the cockpit voice recorder and air traffic control recordings following air accidents.  As illustration, Kikuchi played excerpts of the communication between Capt. Chesley Sullenberger and air traffic controllers during the brief flight that preceded the 2009 Hudson River landing of USAirways Flight 1549. Kikuchi noted stress-induced changes in Capt. Sullenberger's voice that were clear in the recording.

Photo by and courtesy of S. Stefnisson

But if recorded voices reveal clues to the effects of certain "stressors during flight" as Kikuchi believes they do, one has to wonder how else this research can be used to prevent accidents by improving the cockpit environment. It may be possible in the future to use voice recordings to diagnose another persistent problem - pilot fatigue.

And speaking of fatigue, having traveled 12 hours by air and a few hours more by train from Tokyo to my home in Connecticut, I'm going to wrap up this post with a "Sayonara."

After all tomorrow I'm leaving for Iceland and will have more to report later this week.