Canalblog
Editer l'article Suivre ce blog Administration + Créer mon blog
Publicité
La veille c'est déjà demain
27 mars 2015

Artificial Intelligence Could Have Prevented The Germanwings Crash

No level of security screening short of mind-reading could have prevented the crash of Germanwings flight 9525. But what can be done? The New York Times editorial today calls for the American standard that requires two crew members be in the cockpit at all times to be adopted by “all airlines.” This suggestion is reasonable, but would not prevent a team of two pilots from accomplishing a similarly evil deed.

The Times correctly asserts ”Air travel over all remains incredibly safe.” The plane in question, the Airbus A320, has among the world’s best safety records and was the first commercial airliner to have an all-digital fly-by-wire control system. Much of the criticism over the years of these fly-by-wire systems has focused on the problem of pilots becoming too dependent on technology, but these systems could also be a means of preventing future tragedies. In fly-by-wire planes, a story on a previous Airbus crash in Popular Mechanics reports, “The vast majority of the time, the computer operates within what’s known as normal law, which means that the computer will not enact any control movements that would cause the plane to leave its flight envelope. The flight control computer under normal law will not allow an aircraft to stall, aviation experts say.” If autopilot is disconnected or reset, as the New York Times reports it was on the Germanwings plane, it can be switched to alternate law, “a regime with far fewer restrictions on what a pilot can do.”

Germanwings Airbus A320

I just happened to have scheduled an interview with AI pioneer Jeff Hawkins today to talk about the recent upswell of fears about AI and “superintelligence” that he addressed in a post on Re/code. “The Terminator Is Not Coming,” his title announces. “The Future Will Thank Us.” I thought of this story as the news unfolded from the Alps. We are so concerned, it seems, about giving machines too much power that we appear to miss the fact that the largest existential threat to humans is other humans. Such seems to be the case with Germanwings 9525.

Hawkins is the inventor of the Palm Pilot (the first personal digital assistant or PDA) and the Palm Treo (one of the first smartphones). He is also the co-founder, with Donna Dubinsky, of the machine intelligence company Numenta. Grok, the company’s first commercial product, sifts through massive amounts of server activity data on Amazon Web Services (AWS) to identify anomalous patterns of events. This same approach could easily be used to monitor flight data from airplanes and alert ground control in real time of the precise nature of unexpected activity. Numenta open sources its software (see Numenta.org) and is known to DARPA and other government research agencies, so multiple parties could already be at work on such a system.

Hawkins’ approach to machine intelligence, Hierarchical Temporal Memory (HTM), has some distinct advantages over the highly-publicized technique of deep learning (DL). Both use hierarchies of matrices to learn patterns from large data sets. HTM takes its inspiration from biology and uses the layering of neurons in the brain as a model for its architecture. DL is primarily mathematical and projects the abstraction of the brain’s hierarchy to deeper and deeper levels. HTM uses larger matrices and flatter hierarchies to store patterns than DL and the data in these matrices is characterized by sparse distributions. Most important, HTM processes time-based data whereas DL trains mostly on static data sets.

For the emerging Internet of Things (IoT), time-based and real-time data is incredibly important. Systems that can learn continuously from these data streams, like Numenta’s, will be particularly valuable for keeping track of all of those things—including errant airplanes. Could machine intelligence have prevented this tragedy? Hawkins thinks so but notes, “All the intelligence in the world in the cockpit won’t solve any problem if the pilot decides to turn it off.” There will need to be aviation systems “designed for potential override from ground.” What are we the most scared of, individual agency or systematic control? Based on the Germanwings evidence so far, lack of override control from the ground is the greater threat.

I contacted my colleague Dan Reed, who covers aviation and logistics for Forbes.com. He wrote recently on how inexpensive it would be for the airlines to increase their tracking of flights using existing signals. He raised the additional issue of the bandwidth that would be required to control a plane reliably from the ground without significant time delay. This hardware, he says, would require a substantial investment. Securing those transmissions is also important to make sure that the failsafe does not become a backdoor for bad actors. The most important impediment to controlling planes remotely (even temporarily), is philosophical, he says. Even if machines become statistically safer than humans, as Google contends with cars, “how do you prove it would be safer,” Reed asks?

Hawkins sees these fears as a temporary problem, similar to the resistance than many had initially to paying by credit card over the internet. Once there are self-driving cars and self-flying planes, Hawkins says, “people will get over it.” He thinks that events like this aviation tragedy will lead people, “more and more to accept this kind of automated control.” If the software becomes good enough, there will be no need to control it from the ground at all. Anomaly detection and override control software could reside in the aeronautical equivalent of Apple’s secure enclave where it would be inaccessible to the flight crew.

The longer we talked, the more Hawkins warmed to the idea. Through GPS data in Grok, he says, “We could track every plane in the sky automatically with no human intervention, and tell you anytime one of those planes is acting weirdly.… We can find patterns that humans can’t find, and we never get tired.”

“Ultimately,” Hawkins says, “I think we will do that.”

 

Publicité
Publicité
Commentaires
La veille c'est déjà demain
Publicité
Publicité