This paper presents a case study for the use of an IoT-driven digital twin for energy optimization in an automated Surface Mount Technology (SMT) PCB assembly line containing legacy machines. The line was instrumented with multiple sensors for measuring machine-wise activity and energy consumption. A software platform for data aggregation and a discrete-event digital twin of the line were built entirely using open-source tools. Based on the insights gained from data collected over several days, we propose a buffering-based solution for improving the energy efficiency of the line, and evaluate its impact using simulations of the digital twin. The results show that a 2.7x reduction in the energy consumption is possible via buffer insertion without significantly affecting line throughput.