TEMPE, AZ (AZFamily)<\/b> \u2014 Autopilot, autothrottle and autoland have been fixtures in airline cockpits for decades. But those systems follow instructions written in advance. Almost nothing flying today runs on artificial intelligence.<\/p>\n
Students and faculty at Arizona State University are starting to work on changing that.<\/p>\n
A new competition called the AI Challenge for Pilot Assist<\/a> will send faculty-led student teams after some of aviation\u2019s harder problems, including flight planning, emergency response, pilot training alerts and the identification of unusual in-flight situations. It is a collaboration between ASU\u2019s School of Computing and Augmented Intelligence<\/a> and Honeywell Aerospace Technologies<\/a>, which is headquartered in Phoenix.<\/p>\n Applications are open now. Four teams will be selected and will receive up to $10,000 in research funding. Students have until Aug. 14 to apply.<\/p>\n Leading the initiative are Gail-Joon Ahn, an ASU professor of computer science and engineering and a longtime cybersecurity researcher<\/a>, and Jia Zou, an assistant professor of computer science and engineering<\/a> whose work focuses on machine learning and large-scale data management.<\/p>\n Too much data, not enough understanding<\/p>\n On the latest episode of Generation AI, Ahn said the opportunity is in the sheer volume of information a modern aircraft already produces and largely leaves on the table.<\/p>\n \u201cModern aircraft are equipped with numerous sensors that continuously collect and generate massive multimodal data,\u201d Ahn said, listing onboard camera feeds, cockpit audio, air traffic control communications, system alerts, pilot actions, flight telemetry and operational messages.<\/p>\n \u201cIndividually, these data sources provide only partial situational awareness,\u201d he said. Advanced AI techniques are needed to fuse those data streams, Ahn said, producing a fuller picture of an aircraft\u2019s health, its operational status and risks that are still forming.<\/p>\n Ahn pointed to airports where aircraft still experience unexpected collisions on the ground as one place better analysis could help, by pushing predictions and alerts to controllers and pilots alike.<\/p>\n Rethinking the pre-flight safety notices<\/p>\n One target of the challenge is a document pilots deal with before every flight: notices to air missions<\/a>, known as NOTAMs. The advisories flag runway closures, equipment outages, airspace restrictions and other hazards, and they can change while a flight is already underway.<\/p>\n Pilots can miss important notices because they are updated regularly and can change dynamically mid-route, Ahn said.<\/p>\n \u201cIf you can use AI, it can help summarize NOTAMs and even align them with flight data,\u201d Ahn said. \u201cSo eventually you can have real-time alerts. From the pilot\u2019s perspective, they can enhance their decision-making procedures.\u201d<\/p>\n A new target for hackers?<\/p>\n Ahn\u2019s own research field raises an obvious question: does adding AI to a connected cockpit create a fresh opening for attackers?<\/p>\n He acknowledged the risk. Internet-connected devices are susceptible to attack, he said, and commercial aircraft already carry Wi-Fi service.<\/p>\n \u201cThere\u2019s no perfect solution,\u201d Ahn said. \u201cBut as cybersecurity engineers, we have to put cybersecurity principles and solutions in the design phase. We can avoid the potential risks.\u201d<\/p>\n The larger issue right now, he said, is that the data to improve flight safety and pilot performance exist but are not being used. The question is whether it can be turned into early warnings and better decisions while keeping cyber risk down.<\/p>\n Assisting pilots, not replacing them<\/p>\n Ahn emphasized that the goal of the ASU challenge is to assist pilots, not take humans out of the loop.<\/p>\n He also expects the research to say something about the people flying the plane. Linking pilot actions to air traffic control instructions could help catch mistakes and improve human performance in the cockpit, he said.<\/p>\n Beyond the funding, the challenge is meant to give students experience with industry partners in two fields expected to shape the future workforce. It builds on a growing relationship between ASU and Honeywell that includes the Honeywell Innovation Hub on ASU\u2019s Polytechnic campus.<\/p>\n How to apply<\/p>\n Additional milestones, including finalist selection, mentoring sessions and final presentations, are scheduled throughout the 2026-27 academic year.<\/p>\n Watch the full interview on Generation AI, streaming on AZ Family+ and at azfamily.com\/GenAI<\/a>. New episodes post Wednesdays at 7 p.m.<\/p>\n See a spelling or grammatical error in our story? Please click here to report it<\/a>.<\/p>\n