The industry grade smart miniature water bottling plant serves as a comprehensive learning factory for effective rendering of the smart manufacturing concepts end-to-end, integrating PLCs, sensors, actuators, and a Python-based open-source middleware with MySQL for near real-time data (100 milli second) storage and communication. A Flask-based SCADA system ensures zero-latency monitoring and control, promoting remote troubleshooting and robust cybersecurity practices. Students are exposed to the skills that the industry requires, such as remote troubleshooting and predictive maintenance, which require deep understanding of the process as well as how to analyze the data for real-time decision making. This study highlights the effectiveness of learning factories, showing a significant improvement in student performance, with grades increasing from 52% to 89% compared to traditional teaching methods. Additionally, the plant demonstrated the efficacy of predictive maintenance vs reactive and scheduled maintenance in terms of downtime due to failures dropping significantly.