Mining Design Heuristics for Additive Manufacturing Via Eye-Tracking Methods and Hidden Markov Modeling

Abstract

Abstract In this research, we collected eye-tracking data from nine engineering graduate students as they redesigned a traditionally manufactured part for additive manufacturing (AM). Final artifacts were assessed for manufacturability and quality of final design, and design behaviors were captured via the eye-tracking data. Statistical analysis of design behavior duration shows that participants with more than 3 years of industry experience spend significantly less time removing material and revising than those with less experience. Hidden Markov modeling (HMM) analysis of the design behaviors gives insight to the transitions between behaviors through which designers proceed. Findings show that high-performing designers proceeded through four behavioral states, smoothly transitioning between states. In contrast, low-performing designers roughly transitioned between states, with moderate transition probabilities back and forth between multiple states.