New simulator automates air traffic controller training with adapted speech models

Researchers have developed ASTRA, a new simulator designed to train Air Traffic Control Operators (ATCOs) by automating the role of human simpilots. This system addresses the limitations of existing Western-centric speech models, which perform poorly with Singaporean-accented aviation speech, by fine-tuning Automatic Speech Recognition (ASR) models. ASTRA significantly reduces Word Error Rates (WER) to 23.45% and includes an AI-assisted performance evaluation framework for trainee communications. AI

IMPACT This simulator could improve the efficiency and standardization of air traffic controller training, potentially reducing operational costs and instructor workload.

RANK_REASON The cluster describes a new research simulator and its technical methodology, not a product release or significant industry event. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.AI →

• air traffic controller
• ASTRA
• DSPy
• simpilots
• Unsloth

• product
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →