Multimodal learning analytics provides researchers new tools and techniques to capture different types of data from complex learning activities in dynamic learning environments. This paper investigates high-fidelity synchronised multimodal recordings of small groups of learners interacting from diverse sensors that include computer vision, user generated content, and data from the learning objects (like physical computing components or laboratory equipment). We processed and extracted different aspects of the students' interactions to answer the following question: which features of student group work are good predictors of team success in open-ended tasks with physical computing? The answer to the question provides ways to automatically identify the students' performance during the learning activities.