As manufacturers continue to implement AI into company processes, one pharmaceutical-manufacturing leader describes how AI can be applied across operations while facing challenges like data quantity, market shifts and regulatory challenges. 

Helen Clark, the global lead of manufacturing science and technology at Bora Pharmaceuticals, told Automation World how the organization is applying AI to their operations from early product development to commercial manufacturing.

Bora Pharmaceuticals is a contract development and manufacturing organization (CDMO) specializing in formulation development, clinical and commercial manufacturing, and packaging of pharmaceutical products. 

Bora is implementing AI as manufacturers across industries struggle with data quality and quantity and unifying information across company teams. For example, recent research revealed that less than half of manufacturers have fully connected data across design, production, quality and business systems, indicating a gap between operations and what is actually happening on the production floor. 

Clark told Automation World how Bora is facing these challenges through continuing to modernize operations across company systems. 

“Modern manufacturing equipment is essential, but it really only delivers value when it’s supported by the right processes and people,” Clark told Automation World. “That collaboration is what helps create better outcomes. It allows teams to identify issues earlier and build more efficient, reliable manufacturing processes for customers.” 

Responses below are attributed to Helen Clark: 

Automation World: Can you tell me about how Bora manufactures its products, including different product lines and how manufacturing is structured across company teams? 

Helen Clark: Bora's network combines CDMO capabilities with commercial expertise. The company supports biopharma partners through formulation, [chemistry, manufacturing and controls] and regulatory development, scale-up, quality, commercial manufacturing, supply-chain execution and commercialization. Our facilities support multiple product formats, including solid oral dose (pills and capsules), fill finish, liquids and semi-solids, across a range of manufacturing scales.  

An important part of the structure is maintaining continuity as a product moves through the manufacturing process. At sites such as Mississauga, Ontario and Maple Grove, Minnesota, the objective is to keep development and manufacturing expertise connected rather than handing the product from one completely separate team to another. That service-forward model is intended to reduce handoff risk and provide clients with a consistent partner throughout the product lifecycle. 

Bora also uses pilot labs as smaller-scale environments for experimentation and process innovation. At Maple Grove, teams can test formulations and process variables, including how materials flow, compress or behave, before committing time and materials to larger batches. Mississauga also offers GMP pilot-scale capabilities for small clinical or specialized commercial batches (think rare disease with a small patient population), giving teams the flexibility to refine, troubleshoot and validate processes at a smaller scale. 

AW: How has current manufacturing changed compared to the past? What is driving those changes?  

HC: Pharmaceutical manufacturing itself generally changes more slowly. Equipment is a significant investment and may remain in service for decades. In a highly regulated environment, changing a process or piece of equipment can also require extensive revalidation. So, the industry tends to experience incremental improvements rather than frequent, fundamental changes in how products are made.  

Two major factors drive that pace. The first is cost. Pharmaceutical manufacturing equipment requires significant capital investment. The second is regulation. A substantial change to a process or equipment usually requires extensive revalidation for safety, compliance and consistency. This creates a natural preference for proven processes.  

Pharmaceutical regulation is built around repeatability, control and consistent product quality. So while new approaches may offer greater speed or efficiency, they need to be introduced carefully and supported by appropriate validation. And that takes time. 

There are certainly industry trends, like automation, data digitization, and AI analysis, but adoption is not immediate or universal. For most manufacturers, change occurs through targeted improvements versus manufacturing overhauls. 

Just because things move slower doesn’t mean it’s boring work. Today, there’s real advantage in understanding how the industry moves and the importance of flexible manufacturing, which is now less of a buzzword and more of an expectation from pharma partners. Being nimble enough to work on new innovative drug programs is critical and something I love most about my job. 

AW: How is Bora implementing AI in manufacturing processes? 

HC: Bora is actively evaluating how AI can be applied across its operations. Manufacturing generates an enormous amount of data, but historically the industry has not always been able to use that information effectively. One of the clearest opportunities is to make that data easier for scientists and process teams to analyze, identify patterns and eventually support better predictions and decisions for drug candidates and programs.  

AI has been transformational in drug discovery, now the industry is looking hard at how it can streamline manufacturing. For example, by using AI to predict the gap analysis from current processes to the new processes and help with scale up at small scale, tech transfers could be sped up. 

We just announced a partnership with Insilico that will combine Insilico's AI capabilities with Bora’s development, manufacturing, quality, supply-chain and commercialization capabilities. This is an exciting opportunity to design an integrated tool to streamline manufacturing and ultimately reduce batch failures, human error and investigation time. 

AW: What have been some challenges in AI implementation so far? Are there examples of successes vs failures in implementation? If no, why? Does the company have future goals to implement AI? 

HC: One of the clearest opportunities for AI optimization is data. Pharmaceutical companies generate large volumes of information through repetitive development activities, but much of that data is not fully harnessed. There’s a lot of opportunity for AI to help analyze those datasets, identify patterns, generate predictions and improve decision-making. The challenge is making those capabilities reliable and cost-effective for the people who would use them. The technology must also be able to work within a heavily regulated manufacturing environment and produce outputs that teams can trust, with people being the ultimate decisions makers 

The Insilico alliance is an important step in that process, but its scope and operating framework are still being refined. I expect it to be an ongoing work in progress. 

AW: How is Bora using robotics in manufacturing? 

HC: Bora uses robotics and automation in selected operations across its manufacturing network. Examples include automated tube loading, palletization and case packing at several of our sites.  

There are also site-specific applications. The Mississauga facility has robots in operation, Maple Grove has automated or robotic capabilities within warehouse activities, and the Baltimore site uses robotic cleaning tools.  

These technologies are generally introduced where they can make repetitive activities easier to improve consistency or support operational efficiency. As with other manufacturing technologies, broader adoption depends on the business case. 

AW: What is the process of working with clients to ensure that their product needs are met while regulations are adhered to? Can you give an example of a client interaction and the role that Bora plays? 

HC: Customers increasingly expect CDMOs to act as strategic development partners rather than transactional manufacturers. Engagement begins earlier, collaboration is deeper, and scientific guidance throughout development has become a core expectation.  

When a manufacturer is thinking about regulations, it’s important to keep the client’s needs in mind. Clients want to manufacture with the most speed and cost efficiency possible. The goal is to create the most efficient path without compromising compliance. That may include, adapting to existing packaging components, using vendors and suppliers for excipients that are already available rather than creating an entirely new approach. A typical example would be helping a client define the required work upfront, so the scope is technically accurate, and look ahead to potential risks or issues so unexpected requirements do not create delays or additional costs later in the program. 

AW: Regarding regulatory compliance, how do you address challenges of meeting regulations in Canada and the U.S.? 

HC: The target market and regulatory pathway need to be understood at the beginning of a program. Bora can then align the manufacturing process, testing strategy, validation work, packaging and documentation with the requirements of the relevant regulators. 

Our teams look for established methods, existing packaging data and publicly available regulatory information that can support the filing. Using established regulatory information where appropriate can help limit unnecessary development work while maintaining compliance. 

Bora’s partnership with Insilico is intended to eventually use AI to help support regulatory execution as part of a larger, integrated development and manufacturing model, but it’s still in very early stages.