Speaker
Description
The European Space Agency (ESA) has released an AI Harmonisation roadmap, with a focus on "Enabling Artificial Intelligence for Space System Applications." One of the key strategic objectives is to enhance space mission design with the use of AI assistants. Recently, the rapid development of generative AI, has led to significant advancements in automation in various engineering software with the use of AI Agents. Agents use natural language processing via Large Language Models (LLMs) to perform complex reasoning and autonomous interaction with software.
At the same time, there has been significant ongoing work to develop Application Programming Interfaces (APIs) for space thermal analysis tools to support custom workflows. Two notable examples are the ESATAN-TMS Python API which uses a plugin approach and the Thermal Desktop's OpenTD API. This presentation demonstrates initial steps towards using AI Agents to enhance thermal engineering workflows in space thermal analysis tools. Furthermore, the roll of open standards is discussed, particularly with the rise of the Model Context Protocol (MCP) and its use in tool calling.
An initial prototype is presented, merging these two technologies to show an AI Agent taking control of ESATAN-TMS with the engineer in-the-loop to enhance productivity. These examples include parametric geometry construction, electronic unit and honeycomb panel construction, model data access (thermo-optical properties), and automated sensitivity analysis. Several advantages of this approach are highlighted, such as AI enhanced workflows and speed up of repetitive tasks. The approach also permits non-experts to define, access and summarise information from the thermal model through a familiar chat interface.
Finally, some future thoughts are discussed on the use of Agents with respect to testing and verification, non-deterministic results, API development, and data security.
