Dynamic Swarm Intelligence and Generative AI for Integrated Inventive Design
摘要
This paper introduces a novel algorithm that dynamically integrates TRIZ, SCAMPER, and Design Thinking methodologies through generative artificial intelligence (AI) and swarm intelligence. The proposed Dynamic Swarm-Driven Integrated Inventive Design (DS-IID) framework orchestrates a network of AI-driven agents that iteratively explore, refine, and enhance inventive solutions by learning from and inspiring one another. Unlike traditional monolithic or static innovation systems, DS-IID models the co-evolutionary dynamics of distributed design cognition, where agents dynamically select innovation strategies based on real-time evaluations, engage in mutual critique, and integrate valuable solution fragments to evolve the design space. Generative AI amplifies this process by supporting contradiction resolution, creative analogical reasoning, and rapid synthesis of novel configurations. The result is not a mere ensemble of outputs, but an emergent consensus shaped by collaborative reasoning and symbolic adaptation. Experimental validation on real-world innovation challenges confirms that this dialogic swarm interaction accelerates convergence, enhances solution robustness, and expands the semantic reach of inventive exploration. These findings position DS-IID in the innovation management systems that leverage agent-level diversity and dynamic negotiation to enable systematic, explainable, and scalable creativity across complex design contexts.