Evaluation of ideation and problem solving methods for engineering problems

Sammanfattning

An engineer can often face creativity blocks due to constraints put on a project, like manufacturing limitations, time and cost. To support the engineers, ideation methods were developed such as TRIZ and SCAMPER to boost creativity in these contexts. Another ideation tool that has proven to be useful is artificial intelligence (AI). But these methods' effectiveness in engineering tasks is unclear. This thesis compares TRIZ, SCAMPER and AI as ideation tools when applied for engineering tasks. The conventional methods, TRIZ and SCAMPER, are put to test through individual ideation sessions and has 12 engineering students and 6 non-engineers participating as test subjects. Using an AI as an ideation tool was also tested but by the author only. All AI agents and human participants worked on the same design problem, with half of the human participants using TRIZ, the other half using SCAMPER and AI used by the author. Each idea was given an evaluation score and each method an average score for comparison. SCAMPER achieved a higher overall mean score (5.8) when compared to only TRIZ (4.3). But this was driven mainly by superior ratings in helpfulness and ease of use. While TRIZ produced slightly more unique ideas, participants engaged with SCAMPER more readily. The results suggest that between TRIZ and SCAMPER, SCAMPER is the superior choice for quickly learning and applying an ideation method when used for individual ideation for engineering tasks. But when compared with the results from the ideation with an AI, it was clear that neither TRIZ or SCAMPER could outperform this ideation tool. With an idea-quantity score of 9.4 compared to the 2.7 of the conventional methods and an idea-uniqueness score of 7 compared to TRIZ and SCAMPER which got 5.3 and 4.7 respectively, AI has proven to be the most useful ideation tool.