AeroQuery RAG and LLM for Aerospace Query in Designs, Development, Standards, Certifications
Surendra Yadav
Abstract
In the realm of avionics and aerospace, the demand for swift access to critical data is hindered by vast documentation, causing hallucinations, delays, and inefficiencies.To address this issue, our approach leverages the concepts of Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs).Our approach aims to overcome the limitations of LLMs by incorporating real-time data retrieval capabilities through RAG, enabling seamless access to current information. This envisioned chatbot utilizes advanced natural language processing and proactive pattern identification to streamline information retrieval and communication across various aerospace domains.By leveraging advancements in text summarization and utilizing models like Google’s PaLM2, Facebook’s LLaMA, or OpenAI’s GPT-4, we aim to enhance the performance of chatbots in information retrieval. This involves generating training examples and improving text summarization to efficiently address general inquiries related to standards and communication protocols within the aerospace sector but not limited to this.For instance, an aerospace engineer can quickly obtain relevant information on industry standards or communication protocols through the chatbot equipped with RAG, effortlessly taps into external sources to provide up-to-date and relevant information, reducing the need for exhaustive explanations and improving efficiency in information retrieval and communication processes.