Supporting Systems Thinking Application by Data Analysis A Case Study: An Automated Parking System

Abstract

This study applies Systems Thinking (ST) and its tools to define and validate a case study in the early phase of a complex socio-technical research project. We used ST and other tools, including a stakeholder interest map, context diagram, and Customers, Actors, Transformation, Worldview, Owner, and Environment (CATWOE) analysis. These tools are the foundation of ST systemigrams, which is a top-down approach. Further, we support ST with data analysis, which is a bottom-up approach. In this context, we collected and analyzed failure data. We applied machine learning in terms of Natural Language Processing (NLP), Frequent Pattern Growth Algorithm (FBGL) for association rule mining, and the Gensim model to cluster the failure data. The case study indicates that both approaches complement each other as we apply them in an iterative and recursive manner. Data analysis supports ST, and ST guides the data analysis. Furthermore, ST implementation facilitates understanding, communication, and decision-making regarding the case study and its multiple units of analysis. Moreover, we adapt Nonaka and Takeuchi’s model to articulate the tacit knowledge within the Company using Systemigrams, canvas in the form of A3s, and post-its. We adopted the Systems Engineering methodology to construct the canvas we used in the workshops and interviews.