The AI makes aviation safer.

  The AI makes aviation safer. 

MIT created the AI that follows the pilot's behavior. That kind of AI is a so-called technical copilot, whose mission is to assist the pilot. Developers are making similar systems to cars and other transportation tools. The equipment to which AI has access determines its abilities. And the AI copilot is the tool, that can correct the errors that humans make. 

Many crashes happened actually after some other dangerous situation. When pilots first survive some close-call situation, they simply forget to make some routine maneuvers or release the landing gear. In the car, people can stop it and breathe. But in the air, the aircraft continues its flying after those situations. 

And in some other cases human error, like misunderstanding who should fly the aircraft while landing causes an accident. Or the communication between the copilot and captain somehow is misunderstood. 

"With Air-Guardian, a computer program can track where a human pilot is looking (using eye-tracking technology), so it can better understand what the pilot is focusing on. This helps the computer make better decisions that are in line with what the pilot is doing or intending to do. Credit: Alex Shipps/MIT CSAIL via Midjourney" (ScitechDaily.com/MIT’s “Air-Guardian” – AI Copilot Enhances Human Precision for Safer Skies)

In some close-call situations, pilots simply made mistakes and landed aircraft on the wrong airfields or runways. To avoid that kind of situation the airfield or runway must equipped with a communication system that uses laser or coherent radio communication that tells the aircraft the runway that it approaches. 

And then the copilot just pushes the stick, and the captain pulls it. Those situations could avoided by using a system that tells if the pilot and copilot are making opposite movements with control systems. The AI can have the ability to land aircraft at the airfield. If the pilot is unable to operate. 

You can search YouTube by using the words "wrong runway". There is a list of cases where aircraft landed on the wrong runway or wrong airport. It's possible to avoid this kind of case if there is a system that tells what runway the aircraft approaches.  

In some other cases, somebody mistakenly sends two aircraft to the same runway at the same time. And it's possible to avoid those situations by using a system that tells if there is another aircraft on the way to the runway. 

The system requires that all aircraft and flight control just communicate with each other. And then the automatized system just orders to slow rolling speed and wait until the runway is clear. The system requires that all aircraft know the real runway that they approach or roll and share that information with others. 

The need to create a system that can take control of aircraft during hijacking was seen in the 2001 terror strikes where hijackers flew passenger planes towards the WTC building.  If the aircraft is out of course, and doesn't follow flight plans there must be some other way to take control of the aircraft, than just shooting aircraft down. 

The AI must recognize if there is somebody else on the pilot's chair than the pilot. In the case, that the hijacking happens. The automatic system must just fly aircraft to the airfield and call law enforcement to take control of the situation. 

The automatized, AI-controlled autopilots can make biztalk with the airfield when the aircraft is closing. Then the airfield gives the values like closing angle and speed that the aircraft must follow. Same way. Things like wind direction and other important things, like cloud altitude, will sent to the aircraft. 

https://scitechdaily.com/mits-air-guardian-ai-copilot-enhances-human-precision-for-safer-skies/