Automation and Digitalization of Lyophilization

Although, closed loop controls are available, they are not often used. A good understanding of closed loop controls, trust in their use, and reliance on storage and dependability of the stored data are needed.
 
While approaches such as comparative pressure measurement can indicate the end of primary, these methods do not indicate if the critical product temperature (i.e., collapse temperature) was exceeded during processing. In-process data, especially product temperature is needed throughout primary drying to avoid failure of the product. After the formulation and container closure are determined, the product temperature is controlled by manipulating shelf temperature and chamber pressure to ensure product temperature is maintained below a critical temperature. The processing conditions can be optimized at laboratory-scale and are transferred to full-scale, but in-process data are needed to monitor each cycle. Some companies utilize electronic chart recorders (Yokogawa, 2024) to collect in-process data, and some companies collect data via a cloud source for easy comparison of data between batches. There are some companies that have developed methods of collecting product temperature data using wireless sensors that can relay data in real time and would be idea for storing on a cloud base system for future batch to batch comparisons. (Tempris GmbH, 2024) (Ellab A/S, 2024) 

An option that will show immediate benefits with regard to decreasing travel costs, decreasing costs to companies, decreasing carbon emissions, and increasing knowledge in the manufacturing areas is wider implementation of VR/AR.
 

Engineers and operators equipped with VR/AR (as demonstrated in Figure 2) directly interact with supplier companies to solve equipment problems without having the supplier company representative onsite.

​

Another option is the development of digital twins starting with research and development equipment that is well characterized and transferring the knowledge to pilot and full-scale equipment. Some examples of digital twin applications include:

​

1. Implementation of strain gauges on flexible metal tubing that supplies heat transfer fluid to the shelves within a product chamber (Figure 3). The tubing moves every time the shelves are compressed to seal the vials. Over time the tubing becomes weak and develops holes which spray the fluid over the product in the chamber. This can lead to losses of product. Implementing digital twins could alert companies to when these weaknesses were becoming likely.

2. To increase sterility assurance, many manufacturers use isolator technologies with automated loading and unloading systems (ALUS). These systems reduce human intervention and may not allow for the use of temperature probes to monitored the freeze drying process. As a result, conservative parametric cycles are often used with a limited ability to monitor the product temperature during lyophilization. The primary parameters that effect product temperature are shelf temperature, chamber pressure, product resistance and vial heat transfer, which can be modeled. As a result, a digital twin can be created that can follow the process in real-time. The digital twin can provide data to support process deviations and create a design space for the process parameters to support cycle development.

​

The application of digital twins for both equipment preventative maintenance and process design has the potential to increase product quality.