Using motion sensors to understand collaborative interactions in digital fabrication labs

Abstract

Open-ended learning environments such as makerspaces present a unique challenge for instructors. While it is expected that students are given free rein to work on their projects, facilitators have to strike a difficult balance between micromanaging them and letting the community support itself. In this paper, we explore how Kinect sensors can continuously monitor students’ collaborative interactions so that instructors can gain a more comprehensive view of the social dynamics of the space. We employ heatmaps to examine the diversity of student collaborative interactions and Markov transition probabilities to explore the transitions between instances of collaborative interactions. Findings indicate that letting students work on their own promotes the development of technical skills, while working together encourages students to spend more time in the makerspace. This confirms the intuition that successful projects in makerspaces necessitate both individual and group efforts. Furthermore, such aggregation and display of information can aid instructors in uncovering the state of student learning in makerspaces. Identifying the instances and diversity of collaborative interactions affords instructors an early opportunity to identify struggling students and having these data in a near real-time manner opens new doors in terms of making (un)productive behaviors salient, both for teachers and students. We discuss how this work represents a first step toward using intelligent systems to support student learning in makerspaces.