Enhancing concurrent engineering for space mission design with task prioritization and large language models

Space mission design is a challenging problem for several reasons: large and highly non-linear trade spaces, complex and interconnected analyses, and the need to deal with all mission phases (e.g., procurement, integration, qualification, operations, end of-life) with technical, programmatic, and stakeholders’ needs considerations. As opposed to a traditional sequential approach, which can cause late redesigns, missed opportunities, or suboptimal designs, Concurrent Engineering (CE) fosters teamwork and real-time design sessions to simultaneously advance all aspects of a space mission concept. CE facilitates clients’ needs satisfaction through effective trade space exploration, decision-making, and concurrent consideration of all the relevant aspects. This results in valuable, feasible, and consistent designs with a reduced total effort. Argotec is currently implementing a CE framework to improve the efficiency and effectiveness of its processes for mission formulation and design. The implemented methodologies encompass various aspects related to interactions with the client, the organization and planning of sessions, and a retrospective of the study for continuous improvement. During technical iterations, the trade space exploration and point design are supported by the adoption of an Agile approach using story points and task prioritization, ensuring optimal resource allocation across different phases of design. Additionally, given the growing advancement of Natural Language Processing (NLP) techniques across various application domains, this paper also explores the integration of Large Language Models (LLMs) within CE environments. By integrating LLMs, the aim is to optimize systems engineering processes, particularly in information retrieval, a crucial task considering the substantial volume of internal and external documentation involved in the design process. This paper describes the challenges of implementing CE in an industry setting, the methodologies employed to address them, and innovative ideas that Argotec is integrating into its CE framework. These include an Agile approach to CE sessions, with the objectives of task prioritization and activities planning standardization, and the integration of LLMs in the CE process, with the objectives of supporting sessions’ design activities, in particular by leveraging Retrieval-Augmented Generation methodologies to streamline information retrieval processes.