Enhancing experiential learning in data analytics through university–industry collaboration: a pilot study

This study presents a structured university–industry collaboration designed to advance experiential learning in data analytics education. Seven undergraduate students partnered with MineralCom, a leader in the industrial minerals sector, to analyze a decade of de-identified maintenance records (~3 million observations) within a secure Virtual Desktop environment governed by individual non-disclosure agreements. Through the deliberate application of Kolb’s experiential learning cycle, students progressed from cleaning "messy" real-world data to delivering predictive maintenance models. Quantitative pre- and post-course surveys revealed significant perceived competency gains across all domains, with the most substantial growth observed in Data Wrangling (d = 5.65) and Exploratory Data Analysis (d = 4.55). Qualitative reflections underscored a shift in professional identity, as students transitioned from classroom learners to "adaptive experts" capable of navigating domain-specific jargon and ethical data governance. While limited by sample size, this pilot offers a replicable, secure, and resource-efficient framework for bridging the gap between theoretical instruction and authentic industry practice in the AI era.