Multi AI Agents based Black Box Audio Transcriber and Aviation Incident Report Generator

This paper introduces a framework that automates the transcription of black box audio data and the generation of aviation incident reports using AI agents. Traditionally, manual transcription of cockpit recordings is time-consuming and errorprone. The proposed system employs Groq-Whisper for audio transcription and various AI agents for report generation, streamlining the analysis of incident data. Each agent focuses on specific tasks, such as refining transcriptions and structuring conversations, leading to the identification of critical events and causal factors in the reports. This approach enhances accuracy and efficiency in aviation safety investigations, providing a reliable tool for real-time decision-making. The paper highlights the significance of AI integration to improve efficiency while ensuring explainability and trust in aviation contexts, ultimately contributing to enhanced air travel safety.