2026 ASEE Annual Conference & Exposition

Predictive maintenance (PdM) is a cornerstone of smart manufacturing, identifying potential machine failures before they occur. This allows strategic maintenance scheduling, maximizing operational uptime. Despite its importance in industry, traditional engineering education lacks the tools required to give students a hands-on PdM experience. Current educational methods rely on mock PdM datasets, which disconnect students from practical educational experiences built on sensor implementation, real-time data acquisition, and analytics using Machine Learning (ML). To bridge this gap, this research details the development of a low-cost PdM Add-On kit for FrED, a desktop-scale manufacturing system based on industrial fiber draw that serves as an experimental platform for data collection and analysis. The kit, consisting of a modular sensor suite and software, makes hands-on PdM education more accessible. This work aims to demonstrate that FrED, equipped with the PdM Add-On kit, provides an effective and scalable platform for data-centric PdM education. The kit was deployed in a classroom environment, where students collected process data, applied preprocessing and feature extraction methods, and developed ML models. The augmented FrED platform enabled students to engage with the full PdM workflow using a setup mimicking a real-world industrial scenario while growing their advanced analytics skill set. To validate the pedagogical effectiveness of this activity, the study employed a mixed-methods assessment framework grounded in Self-Determination Theory (SDT). Student outcomes were evaluated using three distinct instruments: a technical Concept Inventory (CI) to measure objective learning gains, the Perceived Competence Scale (PCS) to quantify shifts in self-efficacy, and the Intrinsic Motivation Inventory (IMI) to assess engagement and intrinsic motivation.