In the runup to the two-year anniversary of the mysterious disappearance of Malaysia Air Flight 370, a second piece of physical evidence from the plane has turned up off the coast of Mozambique. It’s a triangular sheet about as wide as a spread-out newspaper and half again as long, thought to have come from the skin of the horizontal stabilizer. The first piece of evidence, a chunk of the wing, was found in the Indian Ocean in September.
Surely the most important thing about both pieces of evidence is that MH 370 did indeed crash into the Indian Ocean. The airliner’s last known position was over the South China Sea.
It needn’t have been so hard to keep tabs on a 223-metric-ton- Boeing 777, says Krishma Kavi, a professor of computer science and engineering at the University of North Texas, in Denton. Rather than use—and all too often lose—an on-board “black box” flight data recorder, he says airplanes should transmit data directly to the cloud, through a network of land and satellite-based relays. He dubbed this virtual recorder the “glass box” in a 2009 article for IEEE Spectrum.
‘The fact it took this long to find the debris is again a case for more timely information from the aircraft regarding its location and other flight data,” he said in an email. “Using only satellite information led to miscalculations regarding the potential crash site.”
Today’s black box (it’s actually orange, which shows up better under water) sends out a beacon only for 30 days, which often isn’t enough for salvagers to get a fix on it. International regulators will require that the broadcasting last for 90 days—beginning in 2018. The industry and its regulators are taking their sweet time to implement improvements.
Kavi’s glass box would transmit the data to the cloud—the network of servers that increasingly blankets the earth—and do so in real time.
True, any plan to absorb real-time data from all airliners over the southern seas, the Arctic, and other outlying regions would save only a very few lives. In 2014, the same year that MH370 disappeared, there was one fatality per 2.38 million flights. But that doesn’t mean it’s not worth doing more to understand what’s going on.
“If you think it is important to know what happened and when it happened, it may be used to prevent other accidents,” Kavi says. “And you do not need to do everything we proposed. As we get better with ‘machine learning’ the systems can be designed to learn and decide when it is essential to transmit and when a local decision—on the plane—can be made to determine potential causes and so forth.”
Why not track airliners, too?
Philip E. Ross is a senior editor at IEEE Spectrum. His interests include transportation, energy storage, AI, and the economic aspects of technology. He has a master's degree in international affairs from Columbia University and another, in journalism, from the University of Michigan.