Skip to content Skip to footer

“AutoTRIZ: A Creative AI Instrument that Utilizes Extensive Language Models (LLMs) for the Automation and Improvement of the TRIZ (Innovative Problem-solving) Approach”

The Theory of Inventive Problem Solving (TRIZ) is a widely recognized method of ideation that uses the knowledge derived from a large, ongoing patent database to systematically invent and solve engineering problems. TRIZ is increasingly incorporating various aspects of machine learning and natural language processing to enhance its reasoning process.

Now, researchers from both the Singapore University of Technology and Design and the City University of Hong Kong have introduced AutoTRIZ, a tool that uses Large Language Models (LLM) to automate and improve the TRIZ process. In essence, AutoTRIZ leverages the extensive knowledge and advanced reasoning of Large Language Models to revolutionize design automation and provide interpretable ideation through artificial intelligence.

Specifically, AutoTRIZ can generate solutions for a problem-principle statement provided by the user, which strictly adheres to the TRIZ’s workflow and reasoning process. After the user presents a problem statement, AutoTRIZ uses a four-step reasoning process based on TRIZ principles to generate a well-detailed solution report, outlining the reasoning processes and proposed solutions.

AutoTRIZ makes use of a static knowledge base divided into three TRIZ-related sections to guide the reason control process. As such, AutoTRIZ emphasizes the necessity of controlling the problem-solving process, hence drawing all knowledge related to the problem from the pre-trained large corpora used for training the LLM.

AutoTRIZ underwent an evaluation where its detection results were compared with the analyses provided by human experts from categorized textbooks with full match, half match, and no match scenarios. The results demonstrated that in 7 out of 10 cases, the top 3 detections of AutoTRIZ either fully or partially matched the textbook analyses. As such, it was evident that there was an overlap between human expert results and AutoTRIZ.

In conclusion, the researchers have introduced AutoTRIZ as an artificial ideation tool that uses LLM to automate and improve the TRIZ methodology. With the help of three LLM-based reasoning modules plus a pre-defined function module interacting with the fixed knowledge base, AutoTRIZ has successfully generated interpretable solution reports from problems submitted by the user. From the multiple case studies and quantitative experiments conducted, the effectiveness of AutoTRIZ is quite noticeable, setting the stage for potential future applications in other knowledge-based ideation methods beyond the TRIZ.

Leave a comment

0.0/5