CRMSON AI-powered Crew Resource Management Safety Optimiser Nexus

Flight safety frameworks like Crew Resource Management (CRM) and Threat and Error Management (TEM) assume that flight crews will consistently speak up, share mistakes, ask for help, or admit errors.

New research indicates that pilots often withhold critical safety information when the flight deck environment does not foster a tone conducive to strong team dynamics. This phenomenon, known as “safety silence,” hinders hazard identification and mitigation, weakening aviation safety systems. The findings highlight the critical role of psychological safety in the flight deck. However, research reveals a pilot training gap on this subject – an elevated risk that has remained unaddressed – until now.

Can AI systems enhance interpersonal relationships and strengthen psychological safety by incorporating diverse perspectives in the high-stakes aviation industry?

This project involves creating a dataset to develop the first safety-based multi-agent AI tool designed to build interpersonal capabilities in aviation – the Crew Resource Management Safety Optimizer Nexus (CRMSON). This dataset combines qualitative research focused on capturing “silenced” voices. It uses ethnography, interviews, surveys, co-design and existing open-source incident-reports. The aim is to develop a framework where different AI agents simulate how aviation crews could build relationships and de-escalate conflicts in the plane.

This project will use MDAP’s expertise for data preparation and building machine learning infrastructure capable of integrating data from a variety of sources and support the iterative development of the AI architecture.

Outcomes will involve publications, grant applications and skill transfer, and enable commercialisation options.

CRMSON aims to revolutionise non-technical skills training by leveraging AI-driven technology to provide real-time guidance, microlearning, and data-driven insights tailored to crew dynamics. By integrating academic research with real-world operational experience, it represents the next generation of human performance training, ensuring that aviation professionals develop stronger decision-making, communication, and teamwork skills.

Who's involved

Chief Investigators

A/Prof Fabio Mattioli,Critical Ethnography Lab, Social and Political Sciences, Affiliate ADM+S

Dr Kimberly Perkins, Honorary Fellow, Critical Ethnography Lab, School of Social and Political Sciences

Research team

  • Charlotte Wade
  • Harriet Chard
  • Arielle Meekan
  • Uma Dingemans
  • Chiara Rice
  • Harrison Langdon
  • Nina Zepcan
  • Ridge Pang

MDAP research collaborators

Dr Daniel Russo-Batterham and Geordie Zhang

Project partners

RMIT Aviation Academy

INVOLO