Innovation Machine: Human-LLM Teaming for Aeronautics Research and Discovery

The future of aeronautics will be defined not just by emerging needs and capabilities in aeronautics but by the convergence between emerging areas in aeronautics and other sectors (energy, space, healthcare, agriculture, ecology, etc.). Identifying convergent, cross-sector areas with high transformative potential involves a degree of vision that even transdisciplinary teams can struggle to achieve. In this paper, we offer a set of tools for transdisciplinary teams to collaborate with large language models (LLMs) to co-emerge problem and solution areas across disciplines, sectors, industries, government agencies, and communities. These tools draw from theories of recombinant innovation in the field of innovation management and methods like the heuristic HIT matrix from the field of design theory and methodology. Additionally, we demonstrate new LLM and visualization tools, such as an Innovation Whitespace Generator and Transformative Potential Compass, using a dataset of “idea compost” from the NASA Aeronautics Research Mission Directorate’s (ARMD) Convergent Aeronautics Solutions (CAS) project. We close with limitations and suggested future work around improving these tools and understanding the biases embedded in them (e.g., biasing teams toward problems that advance dominant worldviews and stakeholders). Overall, our goal is to create LLM tools that facilitate collaboration and translation between people who have different expertise and lived experiences so as to enable, not inhibit, the human collaborations needed to identify convergent problems worth solving and transform the future of aeronautics for the benefit of all humanity.