Browse Publications Technical Papers 2023-01-0985

AI Inspired ATC, Based on ANN and Using NLP 2023-01-0985

An Air Traffic Controller(ATC) is a person responsible for the proper Take-Off and Landing of an Aircraft from the runway, and for relaying continuous vital information back and forth from Pilots. The proposed ATC will automate this entire process to reduce human-generated errors and save costs. The entire system will be made using Artificial Intelligence and will use Natural Language Processing and Artificial Neural Networks to create a human-like, but a better-prepared system. The model needed to create the ATC, can be trained on already available crucial flight data. The data must include flight take-off and landing time, along with altered time based on weather, climate and other physical factors. The back-end system of the ATC, can be then made to work on this trained model, and produce correct and calculated flight path and timings for the take-off and Landing. The system will do an automatic Pre-flight checkup, based on weather and other clear-sky conditions, such as birds and overhead flights. If there are no problematic conditions, a flight can be allowed to take-off. Similarly, a flight can be allowed to land, based on a clear runaway and good weather conditions. Also, the system will use Artificial Neural Networks, to pan out an optimized flight path for the aircraft to follow, so as to reach a particular destination by avoiding extra air traffic, and saving fuel.The AI-inspired ATC will also be responsible, to tackle problems faced by pilots based on their requests, voice & mood conditions, which will be processed using a customized NLP Component. Implementing the proposed AI-inspired Air Traffic Controller can significantly reduce errors, save costs, and reduce the overhead of extra time in panic situations.


Subscribers can view annotate, and download all of SAE's content. Learn More »


Members save up to 16% off list price.
Login to see discount.
Special Offer: Download multiple Technical Papers each year? TechSelect is a cost-effective subscription option to select and download 12-100 full-text Technical Papers per year. Find more information here.