Putting the power of agents in your process team’s hands

This demonstration shows digital twins providing instant visibility into batch performance through automated prediction of end-of-batch titer. The dashboard displays critical quality attributes including Viable Cell Density VCD and Titer in real-time, enabling operators to monitor variables and confirm they are trending as expected throughout the batch. When deviations occur, the system visualizes the difference between historical expected paths shown as faded blue lines and current batch performance indicated by yellow and red lines. In the example scenario, VCD and Titer begin deviating at 150 hours, allowing prediction of batch performance issues before they manifest in the actual process. Operators can compare live batches against historical batches, with mean and prediction values displayed across standard deviation ranges. Yellow standard deviation indicates potential batch failure requiring intervention. The system identifies which variables have the largest impact on titer through bar graph visualization, with larger bars indicating stronger impact. This prioritizes which variables to investigate first during deviations. Process recommendations are provided for corrective actions to implement before batch failure occurs. Users can apply and visualize the effect of recommendations before implementing them on the plant floor. The digital twin approach enables proactive prediction and correction of batch deviations to reduce rework, provides earlier insight into batch outcomes, and supports more reliable batch-to-batch comparison.