Tool Object Analysis and ChatGPT Integration for Creative Problem-Solving Algorithm Development

TRIZ, the theory of inventive problem solving, can be used to solve contradictions in high-level problems in inventions and patents. Contradictions arise when two contradictory requirements or conditions cannot be satisfied simultaneously and resolving them can lead to innovative ideas. TRIZ experts have developed 76 standard solutions for applying TRIZ. Though TRIZ’s 76 standard solutions can be applied to a wide range of problem types, they are too complex. We developed a simplified tool object analysis algorithm using TRIZ’s 76 standard solutions. The analysis was used to analyze the interaction between components of a system to identify where problems occur. The interaction between a tool and an object was categorized into beneficial actions (positive relationship), no effect (zero relationship), and harmful actions (negative relationship). When beneficial actions, and gaps or beneficial actions and harmful actions occur in the interaction between a tool and an object, conflicting relationships arise in problem-solving. We classified the interaction between tools and objects into beneficial actions and gaps and beneficial actions and harmful actions to provide appropriate problem-solving strategies for each situation. The analysis has the advantage of specifying the causes and solutions of problem-solving. The analysis was used in creative problem-solving to train ChatGPT. By training ChatGPT for it, specific problem-solving solutions were obtained.