A recurring bottleneck in bioprocess development is the delay between sampling, analysis, and data interpretation. From the process perspective, the key question is: How can data be generated quickly enough to support timely decisions? From the analytical perspective, the challenge is equally clear: How can samples be analyzed more efficiently to generate more data in less time?
In conventional workflows, samples are often collected manually from bioreactors or process vessels, frozen, transferred to analytical laboratories, prepared by specialists, and then analyzed using chromatography or mass spectrometry-based methods. Depending on the assay, sample preparation alone can take several hours. When combined with scheduling, transfer, analysis, and data processing, the overall turnaround time can extend to days or even weeks.
By the time the results are available, the process may already be complete. This limits the ability to react to unexpected trends, adjust process parameters, or use the data to guide development decisions in real time.
PAT is not simply about testing a product at the end of a process. It is a framework for designing, analyzing, and controlling manufacturing through timely measurements of critical quality and performance attributes. The ultimate goal is to ensure product quality by improving process understanding and enabling more effective control.
This represents a shift from reactive analysis to proactive process steering. Instead of waiting for final results after a process has ended, PAT enables teams to collect data during processing and use that information to better understand what is happening as the process evolves.
In this context, automated sampling and analytical integration are essential. They help move analytical workflows closer to the process and reduce the dependency on manual sampling, manual sample preparation, and delayed offline testing.
PAT tools can be viewed as a spectrum based on how closely they are connected to the process.
In a traditional offline workflow, samples are removed from the process and sent to a separate laboratory for analysis. This approach can be powerful and highly detailed, but it often involves multiple handoffs, longer turnaround times, and limited ability to influence the process while it is running.
At-line analysis brings the measurement closer to the process. Instruments may be located near the production or development area, allowing samples to be analyzed more quickly. However, at-line workflows still often require manual sampling, manual transfer, or manual preparation.
Online analysis connects the process directly to an analytical instrument. Samples are automatically diverted from the process to the analyzer and are not returned to the process stream. This enables more frequent measurements and faster data generation, particularly when combined with automated sampling systems.
Inline monitoring represents the highest level of process integration. In this setup, sensors are placed directly in the process stream, enabling continuous measurement without the need to remove samples.
Each step along this spectrum reduces manual intervention and increases the potential for real-time or near-real-time process understanding. Dr. Ayesa’s presentation focused especially on online and at-line workflows using automated sampling, chromatography, and LC-MS-based analysis.
A central element of the presented workflows was the use of the Numera® automated sampling system to connect process streams with analytical instruments. In the described setup, Numera was used to collect samples from upstream and downstream process locations, including perfusion-related process streams and surge tanks.
For online PAT applications, Numera enabled automated sample collection and transfer to a chromatography system. For at-line workflows, it collected and stored samples as retains, which could then be transferred to an automated liquid handling system for more complex preparation and LC-MS analysis.
This flexibility is important because not every analytical method can be fully automated online. Some assays require more extensive preparation, such as purification, enzymatic digestion, or multiple sample-handling steps. By supporting both online and at-line workflows, automated sampling can serve as a bridge between the process and different levels of analytical complexity.
One of the main workflows presented by Dr. Ayesa involved coupling Numera with a two-dimensional chromatography system. In this online PAT setup, Numera collected samples automatically and triggered the chromatography workflow.
For upstream samples, such as harvested cell culture fluid, the workflow used a first-dimension Protein A chromatography step to purify the protein of interest. The sample could then be directed to a second-dimension analytical method, such as size-exclusion chromatography or ion-exchange chromatography, to monitor specific product quality attributes.
This approach makes it possible to follow selected molecular attributes over the duration of a process. In the example discussed, the workflow could support frequent analysis across a long process timeline. Depending on the assay setup, sampling could be performed as often as every one to two hours, with the sampling step itself taking only around 15 minutes and the analytical method taking approximately one to two hours per sample.
For downstream samples, which are typically cleaner and partially purified, the workflow could be adapted for more detailed molecular analysis. For example, an immobilized enzyme reactor could be integrated in the first dimension to support digestion-based workflows, followed by chromatographic separation in the second dimension.
The main advantage is the significant reduction in time from sampling to data readout. Instead of waiting for manual sampling, sample transfer, analyst availability, and offline processing, the system can generate near-real-time analytical information while the process is still running.
While online chromatography is valuable for targeted monitoring, some analytical questions require deeper molecular characterization. For these cases, Dr. Ayesa described an at-line workflow that combines automated sampling with automated sample preparation and LC-MS analysis.
In this workflow, Numera collects and stores samples from the process. These samples are then transferred to a LEAP PAL automation system, which performs preparation steps such as purification and protein digestion. The prepared samples are subsequently analyzed by LC-MS.
This workflow is particularly relevant for methods that would traditionally require a full day of manual work by an analyst. Automated preparation can reduce hands-on time and improve workflow consistency. In the example presented, the system could complete purification and digestion for six samples in approximately six hours, followed by LC-MS analysis.
Although this is not fully online analysis, it still dramatically improves turnaround time. Instead of waiting weeks or months for deeper characterization data, results can become available within a much shorter time frame, potentially within 24 hours depending on the assay.
This is especially useful for monitoring post-translational modifications, glycation, and other product quality attributes over the course of a process. When these data are generated close to real time, they become much more valuable for process understanding and process development.
One of the most important insights from the talk was that online and at-line workflows do not need to compete with each other. Instead, they can be combined to create a more complete process monitoring strategy.
Online chromatography can provide frequent, targeted information on selected attributes during the process. At-line LC-MS workflows can provide deeper characterization data with a short turnaround time. Automated sampling connects both workflows to the process and reduces the dependency on manual sample collection.
Together, these tools help teams better understand what is happening across upstream and downstream process steps. They also support faster identification of critical process attributes and quality trends, enabling better-informed development decisions.
The overall impact of the integrated workflow can be summarized in three major improvements.
First, sampling frequency increases. Manual sampling is replaced or supported by automated sampling, making it possible to collect samples more consistently and more frequently.
Second, turnaround time is reduced. Analytical results that may previously have taken weeks can be generated in days or even hours, depending on the workflow.
Third, data becomes available during the process, not only after the process has ended. This is critical because data are most valuable when they can still influence decision-making.
By combining Numera, Lucullus® software, online chromatography, and automated sample preparation, the presented workflows demonstrate how PAT tools can be integrated into ongoing pilot-plant processes. This moves bioprocess monitoring away from delayed offline analysis and toward near-real-time process understanding.
Dr. Ayesa’s presentation highlighted a practical and highly relevant direction for modern bioprocess development. As the industry continues to move toward Quality by Design, intensified processes, and more data-driven development strategies, the ability to generate timely analytical data will become increasingly important.
Automated sampling and online analytical integration are key enablers of this transition. They help reduce manual intervention, increase data frequency, shorten analytical turnaround times, and provide process teams with information when it matters most.
The result is not only faster analysis. It is a more connected, responsive, and knowledge-driven approach to bioprocess development.
Watch the recorded presentation to learn how automated sampling and online analytical integration can advance PAT in biopharmaceutical process development.