Tuesday, May 21, 2019

Why Is the Pentagon Interested in UFO


"Situational awareness" is the military term for having complete understanding of the environment in which you are operating. A UFO represents a gap in situational awareness. At the moment, when a Navy pilot sees something strange during flight, just about the only thing he or she can do is ask other pilots and air traffic control what they saw in that place at that time. Globally, the number of UFO reportings in a year has peaked at more than 8,000. It's not known how many the military experiences.
Even the most heavily documented incidents end up unresolved, despite interviewing dozens of witnesses and reviewing many written documents, as well as lots of audio and video recordings.
UFOs represent an opportunity for the military to improve its identification processes. At least some of that work could be done in the future by automated systems, and potentially in real time as an incident unfolds. Military vehicles — Humvees, battleships, airplanes and satellites alike — are covered in sensors. It's not just passive devices like radio receivers, video cameras and infrared imagers, but active systems like radar, sonar and lidar. In addition, a military vehicle is rarely alone — vehicles travel in convoys, sail in fleets and fly in formations. Above them all are satellites watching from overhead.

Drawing a complete picture

Sensors can provide a wealth of information on UFOs including range, speed, heading, shape, size and temperature. With so many sensors and so much data, though, it is a challenge to merge the information into something useful. However, the military is stepping up its work on autonomy and artificial intelligence. One possible use of these new technologies could be to combine them to analyze all the many signals as they come in from sensors, separating any observations that it can't identify. In those cases, the system could even assign sensors on nearby vehicles or orbiting satellites to collect additional information in real time. Then it could assemble an even more complete picture.
For the moment, though, people will need to weigh in on what all the data reveal. That's because a key challenge for any successful use of artificial intelligence is building trust or confidence in the system. For example, in a famous experiment by Google scientists, an advanced image recognition algorithm based on artificial intelligence was fooled into wrongly identifying a photo of a panda as a gibbon simply by distorting a small number of the original pixels.
So, until humans understand UFOs better, we won't be able to teach computers about them. In my view, the Navy's new approach to reporting UFO encounters is a good first step. This may eventually lead to a comprehensive, fully integrated approach for object identification involving the fusion of data from many sensors through the application of artificial intelligence and autonomy. Only then will there be fewer and fewer UFOs in the sky — because they won't be unidentified anymore.

Iain Boyd, Professor of Aerospace Engineering, University of Michigan
This article is republished from The Conversation under a Creative Commons license. Read the original article.

No comments:

Post a Comment