In the evolving smart manufacturing world where the importance of downtime reduction is key to operational efficiency, this work defines a setup for an experiment that paves the way for the integration of an Augmented Reality (AR) layer to a Digital Twin (DT) system. This system will attempt to serve as a diagnostic tool to provide operators with an early response advantage while the Digital Twin completes deeper diagnostics. The long-term plan involves comparative analysis of AR’s contribution to this DT system versus a traditional DT interface; this paper focuses solely on the Digital Twin as a baseline. In the case study of a water bottling plant, a digital shadow is implemented to emulate downtime conditions. The architecture incorporates an ESP32 microcontroller, ThingSpeak for relaying data to the cloud, and MATLAB/Simulink for visualizing and modelling faults. A series of tests simulate and document fault conditions, aiming to reveal diagnostic latencies and blind spots that could lead to unexpected downtimes. Initial findings indicate that fault responses within the digital twin interface exhibited significant variability, influenced by both network conditions and fault types. Visual cues emerged at times less than 200 ms, but cloud-driven delays reached more than 14 s, particularly with lag or constrained bandwidth. The result points up the system’s diagnostic capability and susceptibility to interference. This preliminary experiment lays the groundwork for the measurement of future AR incorporation helping fault detection time and cognitive workload in high-pressure manufacturing settings.