Combining TRIZ and NLP to Identify Promising Problem-Solving Approaches for Emerging Technologies: A Preliminary Study

Abstract With the ever-expanding scope and increasing specialization of science research, more novel and niche issues are increasingly being addressed. This invariably leads to subfields developing to the point that they become independent areas of investigation by themselves. An example of such a phenomenon is the emergence of the many subareas stemming from the concept of Industry 4.0. All new such disciplines run into previously seen and unseen issues as they gain prominence and, as a result, practitioners. In this paper, we are focused on the ‘seen’ issues. Problem solving techniques, such as TRIZ, have enjoyed enormous success in engineering practice. The underlying premise behind TRIZ is that there is a small, finite number of methods which can be used to solve any problem and that most such problems have already been encountered and solved possibly in a different setting. Originally intended for physical inventions, TRIZ methods have since been applied to other fields such as patents to great effect. This paper uses natural language processing, machine learning, and abstracts from publicly available scientific articles to find if there exist common problem-solving methods in the engineering research of augmented and virtual reality, digital twins, and additive manufacture. We then present preliminary results and identify avenues for future research.