Project · Assurance

Assurance

Bluebird's third research pillar covers the assurance, explainability, and validation work needed to understand and trust AI agents' decisions.

Research Areas

Formal Agent Assessment

The Machine Basic framework is the world's first formal competency assessment for AI agents in air traffic control, modelled on the NATS basic training course for human trainees.

Assurance & Validation Frameworks

Leveraging the TEA (Trustworthy & Ethical Assurance) platform to build an assurance case and a validation package which provides evidence for the case by measuring accuracy, fidelity and uncertainty.

Ethnographic Studies & Interviews with ATCOs

Study and interviews with human ATCOs to understand how they use tools, work together and make decisions, informing the design of agent interfaces, explainability tools, and human-in-the-loop teaming.

Engaging with Stakeholders

Workshops with stakeholders such as regulators, air traffic controllers, researchers, to ensure our approach anticipates and leads on the emerging regulatory landscape for AI in ATM.

Read Our Assurance Research

Explore publications on assurance, explainability, and trustworthiness in air traffic control.

View Publications