A Digital Twin for Simulated Shipping of Syringes Based on Sensor Fusion and Friction Compensation

Abstract A Digital Twin (DT) is presented for simulating the motion of a rubber syringe seal with hysteretic, nonlinear friction. This model is particularly useful for predicting the effects of high-altitude shipping on the container integrity of pre-filled syringes. The DT is implemented as a virtual sensor network in MATLAB-Simulink, fusing pressure and displacement data with the system dynamics and a PID state observer for friction compensation. The PID state observer represents a simple and accurate approach when the key state (position) is measurable using high-accuracy, low-noise sensors and external forces are well-defined. Performance was qualitatively assessed by comparing virtually sensed friction with known tribological characteristics of the stopper. The PID state observer successfully captured nonlinear friction behaviors, including nonlocal hysteresis, break-away forces, and lubrication effects. Predictive accuracy was tested using k-fold cross-validation across syringe populations with varying lubrications conditions, achieving errors on the order of tenths of a millimeter. A key limitation is that the predictive capability is restricted to the specific external force conditions present during sensor data collection. However, two case studies demonstrate the DT's practical application in assessing real-world shipping processes. Through realistic simulation, the DT provides a valuable tool for ensuring container integrity during high-altitude shipping of pre-filled syringes.