CSL Behring's Robotics Push Is One Of The Biggest We've Seen Yet
A conversation with Anthony Kaye, CSL Behring
CSL Behring’s Project Aurora expansion build-out at the company’s Broadmeadows campus in Victoria, Australia, is drawing industry attention for its use of process simulation, integration with the company’s global systems, and a dizzying number of robotic and automated features.
CSL is this year’s 2025 Pharma 4.0 category winner of the International Society for Pharmaceutical Engineering’s (ISPE) Facility of the Year Awards (FOYA) program. The award recognizes companies that implement digital innovations and change to keep pace with advancing technology.
CSL and other FOYA category winners will be recognized at the 2025 ISPE FOYA Banquet and Awards Celebration, commemorating FOYA’s 20th anniversary, which takes place at the 2025 ISPE Annual Meeting & Expo.
At the Broadmeadows site, Project Aurora expanded CSL’s plasma fractionation capabilities nine times from the existing capacity. As a result, this supported CSL’s ability to supply increased global volumes of immunoglobulin and albumin therapies. To help us understand what it takes to see that kind of improvement, the company’s Anthony Kaye, senior director of base fractionation, offered to answer some questions ahead of the conference. Here’s what he told us.
Robotics and the sheer number of automated solutions obviously played a central role in the FOYA win. Can you walk through the technologies you chose and why? How has this level of automation changed your workforce dynamics?
Kaye: CSL’s Project Aurora, particularly Facility F, showcases a sophisticated integration of automation and robotics that was instrumental in securing the FOYA 2025 recognition. The facility employs a wide array of automated technologies, including robotic systems for plasma pooling, automated cleaning, powder handling, and waste disposal.
These systems were chosen to improve safety, enhance operational efficiency and reduce process losses. For instance, automated plasma debottling and debagging machines streamline the initial stages of processing, while the automated clean-out-of-place (COP) robotic filter press machines minimize manual handling and reduce contamination risks.
The facility also features a fully integrated warehouse system supported by automated guided vehicles (AGVs), which navigate a dedicated corridor loop to deliver materials on demand. This not only improves material flow but also enhances safety with the separation of material pathways from pedestrian walkways.
Additionally, the use of Siemens PCS-7 for process control and Korber PAS-X MES for batch record management has digitized operations, significantly reducing batch review times and human error.
This high level of automation has reshaped workforce dynamics at CSL. The transition from manual to automated processes necessitated a shift in workforce capabilities. CSL responded by implementing comprehensive training and development programs to upskill employees in digital tools and automated systems. Engineers involved in the design and commissioning phases transitioned into operational roles, ensuring deep process knowledge was retained. This approach has improved retention by offering career development opportunities and fostering a culture of continuous improvement. The emphasis on safety, efficiency, and innovation has also made CSL an attractive employer in the advanced manufacturing sector.
You mention using digital twins for process simulation. How exactly are you using these models today in process optimization? Have you seen measurable improvements in throughput, quality, or issue resolution?
Kaye: CSL’s Facility F under Project Aurora leverages digital twin technologies to optimize manufacturing processes and reduce operational risks.
A virtual replica of the facility’s automation systems, powered by Siemens’ SIMIT platform, is currently in operation. This allows CSL to simulate, assess, and refine process changes before implementing them in the physical plant. By modelling the automation system, the team can test various scenarios, identify bottlenecks, and validate control logic without interrupting live operations. The SIMIT platform also supports training by allowing operators to interact with simulated environments, improving familiarity with systems and reducing onboarding time.
In addition, CSL also conducted a highly detailed dynamic process simulation early in the project using the INOSIM platform, covering all process and cleaning steps. These simulations enabled the identification and prioritization of bottlenecks and informed continuous improvement initiatives. Additionally, they helped optimize staffing models by analyzing operational flows and resource requirements.
The measurable benefits of using digital twin technologies at Facility F include observed improvements in productivity due to streamlined workflows and reduced downtime. Furthermore, the ability to simulate changes before deployment has led to a reduction in errors and deviations, improving issue resolution and overall reliability. This proactive approach to process management exemplifies how digital twins can transform pharmaceutical manufacturing by combining precision, agility, and foresight.
What lessons from selecting, building, and rolling out your MES might help other companies in similar situations? How did you overcome challenges related to integrating the MES with existing knowledge or data infrastructure?
Kaye: CSL’s selection of the MES platform in Facility F was the result of an extensive study driven by a global base fractionation project. This global project leveraged all expertise in the CSL network and assessed different MES options across three regions (i.e., Europe, Australia, and the U.S.), according to the best options in the market at the time. At that point in time the Korber PAS-X MES solution was adopted. The key lesson learned is to capitalize on the knowledge and experience of the SMEs across the business network when selecting critical manufacturing systems. The selection criteria should not be limited to the scope of the project; they should align with the company’s digital strategy and the broader goals of operational efficiency and compliance. Furthermore, the MES technology continues to improve and advance year by year. Any new company getting into MES design should analyze the market at the time of implementation.
The PAS-X MES solution was chosen due to CSL’s past experiences and existing plants. It allows digitizing batch records, automating workflows, and integrating with other systems such as ERP, LIMS, and PCS.
We overcame challenges related to integrating the MES by actively involving operations SMEs throughout the MES workflow design process. Their deep process knowledge was critical in translating manual procedures into digital workflows, identifying key data touchpoints, and ensuring that system logic aligned with real-world operations. This collaborative approach helped bridge gaps between existing infrastructure and MES capabilities, resulting in a more intuitive and effective solution. Furthermore, CSL’s PAS-X MES needed to be customized to support advanced data exchange to PCS, as per the local request, and this enables real-time monitoring and control of production processes. This functionality was not part of the off-the-shelf product but we managed to work with the vendor to provide a suitable solution.
Did you deploy edge computing for any real-time control functions, or is your architecture primarily cloud-based? What led to those decisions?
Kaye: CSL’s Facility F under Project Aurora primarily relies on a robust and secure on-premises IT infrastructure rather than a cloud-based architecture.
At the time Project Aurora was launched in 2017, cloud-based technologies were not yet a proven or widely accepted solution for process control in the pharmaceutical industry. Regulatory uncertainty, data integrity concerns, and latency risks made cloud adoption for GMP-critical operations a significant challenge.
To uphold CSL’s patient-focus value and avoid introducing unnecessary risk to product quality and compliance, the team opted for traditional edge computing systems. The facility uses Siemens PCS-7 for process control and Korber PAS-X MES for manufacturing execution, both of which are integrated into a multilayered, redundant IT environment designed to ensure high availability and minimal downtime. This infrastructure supports real-time control functions locally, which is critical in pharmaceutical manufacturing where minimal latency, reliability, and regulatory compliance are paramount.
This decision reflected a conservative yet strategic approach, prioritizing robustness, reliability, and regulatory alignment over emerging but unproven technologies. As cloud maturity has evolved, future phases of Facility F and other CSL’s facilities may revisit hybrid architectures.
Did you take a modular design approach to any of the facility systems (utilities, skids, automation cells)? What parts of the build benefited most from modularization — and what didn’t?
Kaye: Yes, CSL’s Facility F was designed with a strong emphasis on modularization, which proved to be a cornerstone of Project Aurora’s success. The facility comprises three distinct production modules, each capable of independent operation and maintenance. This modular design enabled continuous production while allowing staged construction and commissioning, significantly accelerating the project timeline. For example, Module I was fully commissioned within 36 months, a pace 25% faster than comparable projects.
Modularization extended beyond the cleanroom areas to include utilities and automation systems. Technical corridors were introduced to house up to 90% of valves, pumps, instrumentation, and pipework outside of graded cleanroom zones. This design simplified cleanroom layouts, improved maintenance access, and reduced HVAC complexity and energy consumption. Additionally, process and utility equipment were prefabricated off-site as skid-based units, allowing parallel construction and reducing installation time. Electrical cabinets were designed with multi-tier configurations to enable safe power-up and concurrent termination work.
While modularization brought substantial benefits in terms of speed, scalability, and operational flexibility, certain areas posed challenges. For instance, integrating the third module for dual-mode operation — supporting both plasma fractionation and novel material production — required extensive design coordination, clash detection, and simulation. Although successful, this adaptation was more complex than standard modular deployment. Nonetheless, the overall modular strategy enabled CSL to future-proof the facility, accommodate evolving product pipelines, and maintain uninterrupted operations during expansion or maintenance.
Have you integrated predictive maintenance or AI-driven alerts into facility or process equipment monitoring? Can you share any early ROI or reliability gains?
Kaye: No, we have not yet integrated predictive maintenance or AI-driven alerts into our facility or process equipment monitoring. However, we recognize the potential value these technologies offer and are actively exploring opportunities for future integration. Our aim is to further enhance reliability and operational efficiency of Facility F, and we anticipate that implementing such solutions will contribute positively to long-term performance outcomes.
About The Expert:
Anthony Kaye is the senior director of base fractionation at CSL Behring, where he has worked since 2015. Previously, he was the general manager at Boron Molecular and before that, a manager at IDT Australia Ltd.