Capacity to capability in U.S. pharma manufacturing

By August 2026, pharmaceutical companies had announced roughly 500 billion USD in U.S. investments across research, development and manufacturing. But building a facility is not the same as being ready to produce. Turning new capacity into dependable manufacturing capability requires aligning therapy and process development with equipment, supporting infrastructure and operations. The aim is to produce high-quality medicines at the required scale, where patients need them. 

Manufacturers must also respond to supply-chain vulnerabilities, cost pressures and constraints on skilled personnel while meeting regulatory requirements. Adding capacity does not, by itself, resolve those pressures. Increasingly complex therapies, uncertain clinical outcomes and evolving demand make it harder to translate scientific innovation into reliable supply. Manufacturers must manage that uncertainty from clinical development through commercialization. 

That capability begins before the first commercial batch. A process that supplies early clinical trials must evolve to support later-stage studies and commercial demand. Manufacturers need to optimize operating conditions across different volumes, equipment configurations and batch sizes while preserving the therapy’s intended quality attributes. Teams must develop the therapy and its process together and resolve uncertainty early. They must also retain the knowledge needed to transfer and reproduce the process as requirements change. 

Reaching stable production sooner helps manufacturers support clinical programs and product launches and realize the value expected from their investment. Doing so cost-effectively requires coordination across two complementary lifecycles. One covers development of the therapy and its production process. The other covers the manufacturing capabilities needed to produce it, from design through operation. Product quality requirements guide process design. Process understanding informs engineering, and production experience strengthens subsequent development decisions. 

Explore ways to improve production readiness

Assess pharmaceutical manufacturing readiness beyond construction 

The work begins with how manufacturers assess progress toward production. A project dashboard might show a site is “85 percent complete,” but that percentage alone does not establish operational readiness. An unresolved question about process controls or utility performance can matter more for start-up than the percentage suggests. 

The question is how much uncertainty remains and whether it could delay clinical supply or a product launch. An expansion or a transfer to another site requires the same scrutiny. The process must work within the capabilities and constraints of the receiving operation. The consequences are commercial as well as technical. Unresolved uncertainty can delay supply, increase the work needed to qualify the process and leave expensive manufacturing assets underused. 

Troubleshooting a process disruption requires collaboration among manufacturing science and technology, process development, quality, engineering and facility teams. Those relationships are equally important during development and before production begins. Scientific evidence, process definitions and engineering information may pass through disconnected documents and systems. The reasoning behind a formulation choice or process parameter can be lost along the way. A technology-transfer package may describe what to do without preserving why it matters or which quality attributes it protects. It may also leave receiving teams unsure how their equipment can reproduce the required conditions. 

Digital twins can help teams evaluate process behavior and equipment assumptions before building or modifying a physical system. A continuous digital thread can preserve relationships between scientific evidence, material attributes, process parameters, equipment capabilities and quality outcomes. Teams can trace these relationships through development, technology transfer and production. Models, experimental data and the reasoning behind development decisions can move with the process from laboratory studies through clinical supply to commercial production. Receiving teams can assess which conditions must remain consistent and which can change, reducing repeated experiments and avoidable technology-transfer effort. Operating results can then inform further development and engineering. Identifying gaps early can reduce redesign, repeated testing and the work needed to qualify the process, while decisions are still easier to change. 

Resolve process uncertainty across clinical development and scale-up 

Scale-up puts process understanding to the test throughout development. A mixing regimen that works in a small vessel may perform differently at pilot or commercial scale. Fluid flow, heat transfer and mass transfer change with equipment scale and geometry. Teams must understand those effects as batch sizes increase between clinical stages. The same applies when different equipment or sites are needed to meet supply requirements. 

Timing matters. Evaluating alternative volumes, operating conditions and equipment configurations during development gives teams more room to adjust. Discovering a mismatch after installation can require changes across equipment, software, testing and qualification documentation. The technical issue may be the same, but missed quality or yield targets now carry greater business consequences. Late discovery can consume scarce technical resources, disrupt clinical supply and delay commercial-scale manufacturing. 

Modeling and simulation move consequential learning earlier in the lifecycle. Process models address chemical and biological questions, while fluid dynamics help explain equipment-scale behavior. Virtual commissioning tests automation logic against simulated equipment or processes before the physical system is available. These approaches allow teams to compare production options and investigate operating conditions at different volumes. Their usefulness depends on whether each model adequately represents its intended application. As experiments and production runs add evidence, teams can refine the models. This helps them focus physical testing, optimize the process and progress toward scaled production. 

J&J Innovative Medicine’s chemical development teams in Belgium used a digital process twin to evaluate a new process before scale-up. The approach worked at lab scale, but the relevant parameters changed when the team considered industrial scale. The model predicted that the proposed configuration would not work at that scale. The team could avoid further physical trials of that configuration and focus development resources on more promising alternatives. Digital evidence does not replace commercial-scale qualification and process validation. 

The PreCheck Pilot Program from the U.S. Food and Drug Administration (FDA) reflects the value of earlier engagement. It engages selected manufacturers during facility development and pre-production at new U.S. facilities; extensions of existing facilities are excluded. Through pre-operational reviews, FDA evaluates facility design and readiness information and provides early technical guidance. The goal is to identify facility and manufacturing-readiness issues early and reduce the risk of downstream delays. The program also aims to support timely facility assessment and inspection within the application review process. While participation does not guarantee approval or inspection outcomes, early engagement can help clarify expectations, support a more efficient review process and reduce avoidable delays. 

Learn how simulation supports scale-up

Align manufacturing infrastructure with process and supply requirements 

Understanding the process is one part of establishing readiness. Delivering the required quantity and quality also depends on how production equipment, material flows and supporting infrastructure work together. Those capabilities should follow the therapy’s process and supply requirements. These include the batch sizes, production locations and credible future uses the operation must support. 

Roche’s Apollo 11 greenfield pharmaceutical manufacturing project in North Carolina illustrates this approach. Working with us and other technology partners, Roche is using digital twins and manufacturing simulation to evaluate design decisions. The team began by comparing two layouts and simulating material and personnel flows before implementing either layout. Comparing these alternatives allowed the team to make design changes before they became more expensive and disruptive. While actual performance must still be verified during commissioning, evaluating these alternatives digitally gives teams a stronger basis for design decisions and helps identify potential constraints before construction and installation make changes more costly. 

Capacity discussions often focus on the production line, assuming the surrounding facility will meet its requirements. Expansions can test that assumption. Depending on the process, a new line adds demand for purified water, steam, cooling and electrical capacity. These requirements need to be assessed alongside the production plan. Teams must establish whether different operations will compete for the same utility at the same time. Installed equipment represents potential capacity, but only usable capacity creates value. Utility constraints or inadequate infrastructure can limit output even when the production equipment itself is available. 

The facility must also maintain the environmental conditions required by the process. Deviations in temperature, humidity or pressure can affect production and trigger an investigation. Critical environments, including cleanrooms and laboratories, require monitoring, controls and records appropriate to their intended use. Infrastructure is part of the manufacturing environment, and its performance must be understood in relation to the process it supports. 

This makes energy performance an operating concern as well as an environmental one. Monitoring consumption against production demand can help teams identify inefficient equipment or avoidable utility use. Reducing avoidable energy and utility use can lower operating costs, provided that improvements maintain the conditions required for product quality. 

Explore infrastructure for reliable production 

Use production data to improve pharmaceutical manufacturing 

Once production begins, operating data can help teams test development assumptions against the actual performance of the process and facility. By batch 100, the organization should understand its process better than it did after batch one. That learning should translate into more consistent output, improved yield and less effort spent investigating and reviewing deviations. Each run should add to what the organization knows about the process. 

A temperature reading or pressure setpoint needs a link to its batch, process step, material and applicable process definition. That context helps teams compare quality outcomes with development assumptions and understand the effects of material fluctuations, equipment wear and environmental shifts. It also supports Industrial AI: artificial intelligence applied to tasks such as analyzing process variation or detecting equipment anomalies. Preserving these relationships makes production experience useful for subsequent process optimization, technology transfer and scale-up decisions. 

At AGC Pharma Chemicals’ active pharmaceutical ingredient site in Barcelona, teams integrated the control system with manufacturing execution and electronic batch records. This removed more than 530 manual operator entries. The result was more complete digital documentation of its production steps and less information for employees to transcribe. For manufacturers facing workforce constraints, this can reduce repetitive work while retaining skilled people’s responsibility for process and quality decisions. 

A more complete record also creates a better basis for analysis. Review by exception can direct attention to departures from defined requirements. Process analytical technology can combine measurements and models to assess relevant quality attributes during production. Equipment-condition monitoring can identify developing problems for maintenance teams to investigate. These applications use information differently. Connecting systems does not make every insight an instruction to change a process or release a batch. These applications can reduce review effort and direct technical resources toward actual exceptions. They can also help identify developing equipment problems before they cause more costly disruption.

GSK’s work with Siemens and Atos illustrates the additional steps between measurement and action. In a proof of concept for vaccine-adjuvant manufacturing, a digital twin used process data and models to estimate particle size and calculate flow-rate adjustments. Laboratory tests demonstrated direct control in response to introduced disturbances. In a subsequent small-scale engineering run in a good manufacturing practice (GMP) facility, the twin recommended adjustments to an operator. It did not implement them directly. 

Applying such an approach in commercial manufacturing requires models and controls that are fit for their intended use. Appropriate validation and change control remain necessary. Its value should be visible in outcomes such as fewer deviations, more reliable output, less review effort, lower downtime and reduced quality-related losses. 

Industrial AI needs the same discipline. Contextualized, traceable data can help teams investigate deviations, analyze process performance or identify equipment problems earlier. A shared, governed context connects relevant scientific and operational evidence while each discipline retains its specialist systems and responsibilities. Each application needs a defined purpose and evidence that its outputs are reliable enough for the decisions they support. Its value should be tied to a measurable outcome, such as shortening investigations or preventing disruption. 

See how production data can improve operations 

Reuse process knowledge as manufacturing requirements change 

The knowledge accumulated during development and production also provides a starting point when requirements change. A manufacturer may need a different clinical batch size or a new product. It may need to transfer to another site or manufacturing partner, or increase output while maintaining existing supply. A governed process definition, supported by development history and quality requirements, helps teams assess what can be reused and what must change. This can reduce the effort required to recreate process knowledge and make each transfer or modification a more informed engineering exercise.

At Bayer’s aspirin facility in Bitterfeld-Wolfen, Germany, teams introduced new manufacturing execution and automation systems in stages while production continued at the site. Connecting manufacturing execution and process automation allowed subprocesses to be configured without duplicating the modeling work in separate systems. For this project, Siemens reports up to an 80-percent reduction in engineering effort for subprocess modeling. The comparison was with separately modeling the manufacturing execution and control systems. The new approach supported paperless manufacturing, electronic batch documentation and faster implementation of recipe or automation changes. 

This is one practical meaning of adaptability: reducing the time, cost and operational risk involved in a defined change. Standardized equipment interfaces and reusable automation structures can also reduce custom integration when compatible modules are added or replaced. They do not make a facility suitable for every modality. Different therapies may require different containment, cleaning, environmental conditions or process equipment. Flexibility has to be designed around credible future uses. Qualification and validation requirements must be considered from the outset. This helps manufacturers avoid unnecessary complexity while investing in the adaptability most likely to support future products.  

See how modular automation simplifies change 

Connect therapy development with manufacturing capability 

Making those changes manageable requires continuity across research, formulation, process development, technology transfer, quality, supply planning, engineering and production. A change in the therapy’s quality requirements or formulation can affect process parameters, equipment selection, automation and utility demand. Evaluating those consequences together helps teams make earlier trade-offs that protect clinical supply, launch schedules and expected production performance. Manufacturing evidence must also feed back into process understanding, so that knowledge improves across both lifecycles. 

Manufacturers should look for technology partners that support these handoffs across development stages, manufacturing sites and specialist systems. Pharmaceutical operations use varied technologies, so partners must be able to work with existing equipment and other technology providers. Open standards and interoperability should support secure information exchange. Responsibilities for interfaces, data quality and the context needed to interpret information must be clear. Traceability between requirements, process definitions and execution helps teams assess change and reuse established knowledge without unnecessary custom integration. 

We work across several of the points where those decisions meet. Our work spans process and production simulation, engineering, quality, automation, manufacturing execution and supporting infrastructure. The opportunity is to preserve context as decisions move from process understanding into engineering and production, then use operating evidence to inform further development. This can reduce the effort required to recreate information across lifecycle stages and make future changes easier to implement. Connecting scientific research and development, process modeling, enterprise recipe management and manufacturing execution can help teams carry controlled process definitions into production. Their supporting evidence can then remain available as teams reuse that knowledge across scales and sites. 

For current investment decisions, four actions deserve attention: 

• Develop the therapy and its process together, testing how operating conditions change across clinical and commercial scales. 
• Assess equipment, utilities and environmental controls against the production requirements they must support. 
• Preserve process definitions and supporting evidence so receiving teams can assess changes across sites and scales. 
• Connect production data to its batch and process context, then use operating results to improve development and engineering decisions. 

Industrial AI can extend this learning by helping teams analyze process variation and identify equipment problems earlier. The next step is to select a defined manufacturing problem, establish the necessary data and test whether an application improves a measurable outcome. Those decisions belong in today’s investment plans, while teams can still shape the systems and workflows they will depend on.   

Explore Siemens pharma manufacturing solutions 

Published: October 30, 2026