Researchers are attempting to understand how cognitive load impacts an aviator's ability and decision-making and using AI to develop improved training.
Researchers at UNSW Canberra are developing a revolutionary new training method for aviators that will leverage AI to create a smarter, more adaptable program aimed at improving performance and safety.
The research team, led by Dr Oleksandra Molloy in partnership with CAE Australia, have uncovered critical insights into how cognitive load affects human performance and situational awareness while completing complex activities.
They did this primarily through a series of simulated missions completed by UNSW Canberra students, where they tracked cognitive load via physiological responses such as heart rate variability and situational awareness by utilising cutting-edge eye-tracking technology.
Dr Molloy said that training for mission aviators had not changed for decades and that the study would provide valuable insight for the Australian Defence Force. The project has received grant funding through the Defence Trailblazer program.
“This project provides an opportunity to bring new and innovative approaches to mission aviator training,” Dr Molloy said.
“By exploring the possibilities of adaptive training systems generated with the help of AI, training can be tailored to specific missions and allow trainers to determine what aspects require the most attention.
“What we learn from this project could also have implications for training people in other high-risk, high-stress settings where they’re required to complete complex tasks.”
Dr Molloy said the study confirmed a long-standing psychological principle known as the Yerkes-Dodson Law, which suggests that performance peaks under moderate levels of cognitive challenge.
“The results demonstrated that the best performance was achieved while performing tasks under medium task load (95% accuracy), while performance slightly declined under both high (91.11%) and low (87%) load conditions,” she said.
“These results suggest that too little challenge may lead to disengagement, while excessive demands can strain attention and working memory.”
In addition to the objective measures such as heart rate and skin temperatures, participants were asked to report how they perceived the demands on their mental capacity using an established tool, the NASA Task Load Index.
While heart rate remained relatively stable across participants, heart rate variability - a more sensitive measure of cognitive effort - decreased significantly as load increased. This reduction in variability, particularly between low and medium load conditions, aligns with what is known of the sympathetic nervous system’s activation during heightened mental effort.
Situational awareness, a critical factor in aviation safety, was found to decline significantly under high cognitive load. Dr Molloy said this was likely due to limitations in participants’ attention or information overload, which impacts a person’s ability to perceive risks.
CAE Australia Industry Partner Investigator, Dr Gary Eves, said mission aviators could improve their situational awareness by incorporating the company’s simulators during training.
“Situational awareness isn't just a skill, it's the foundation of effective decision making and mission success,” Dr Eves said.
“Using the CAE UAS Mission Trainer and biometric measures, we've demonstrated the value of immersive training environments in revealing critical insights into cognitive load and performance.
“With advanced technology and systems, training can be delivered to the individual's needs in real time and provide instructors with critical performance insights, helping future aviators to anticipate, adapt, and respond decisively.”
With the first stage of the project complete, the research team are taking what they’ve learned from the mission simulations and applying it to the development of new training methods.
“This is a significant step towards adaptive training systems that respond to an aviator’s cognitive state in real time,” Dr Molloy said.
“By integrating physiological monitoring with AI-driven feedback, future training programs could dynamically adjust task difficulty, pacing, and instructional support to maintain optimal cognitive load and situational awareness in participants.
“As aviation continues to evolve with increasing automation and complexity, these findings offer a roadmap for designing smarter, safer, and more effective training environments - especially for those just beginning their journey in the skies.”
