Robert Wethman, Associate Scientific Director, has spent more than two decades at Bristol Myers Squibb working at the intersection of analytical science and manufacturing, where the discipline plays both a foundational and increasingly strategic role in pharmaceutical development.
In this installment of Frontline Pharma, Wethman reflects on the growing complexity of modern therapeutics and the pressure it places on analytical workflows. From the promise of real-time spectroscopy to the rise of AI-driven modeling, he argues that the future of biopharma will hinge not just on better tools, but on how effectively they are implemented – requiring early alignment, operational insight, and a clear return on investment.
You’ve spent much of your career in industry rather than academia. What drew you to that path, and what advice would you give young analytical scientists weighing those options today?
A career in industry offered a wide range of pathways and the opportunity to apply my academic training in different ways as new challenges emerged. That variety was a strong draw, and it has allowed me to adapt my focus as technologies, molecules, and manufacturing approaches have evolved.
For early-career analytical scientists, my advice is not to define your path too narrowly. You need to be an active driver of your career, open to exploring opportunities as they arise. The industry is constantly changing, and long-term success depends on being willing – and able – to adapt alongside it.
From your vantage point, how is analytical science viewed within large pharmaceutical organizations?
At its core, analytical science supports manufacturing by delivering robust, reliable analytical methods. At the same time, it is increasingly a strategic driver, because decisions around how that analytical support is enabled – what technologies are used, when they are deployed, and how they are integrated into development and manufacturing – have a direct impact on timelines and outcomes.
While the ultimate goal is always accurate and reliable measurement, the pathways to achieving that goal vary widely. Aligning analytical strategies with overall development and manufacturing strategy is critical, particularly as timelines compress and processes grow more complex. Real-time spectroscopic monitoring, for example, can add value across the lifecycle, but its adoption requires careful evaluation of cost versus benefit at each stage.
What are the biggest analytical bottlenecks you see today across the drug development and manufacturing pipeline?
The biggest challenge is the increasing complexity of molecules, modalities, and manufacturing processes. Any one of these factors can slow development, but in reality, all three are advancing simultaneously – while timelines continue to compress.
This combination creates natural bottlenecks. One way analytical science can help mitigate them is by increasing the speed of analysis. Spectroscopic methods offer a clear advantage in this area, and when appropriate, they provide an effective lever for keeping pace with growing complexity.
What have been the hardest practical challenges in implementing spectroscopy for real-time process monitoring?
Successful implementation depends on early knowledge sharing and a clear, shared understanding of goals, limitations, and long-term support requirements. In many ways, the challenges are similar to any analytical method transfer, but with a much stronger emphasis on collaboration with operations.
Traditional method transfers typically occur from lab to lab. Real-time spectroscopic methods, by contrast, must account for operational impact. Time invested upfront to understand those impacts is essential to realizing the full benefit. The most persistent challenges have been delivering high-quality data that meaningfully reduces cycle time while minimizing ongoing operational support. My early experience in operations has been invaluable in anticipating and addressing these realities.
How close are we to routine, truly continuous bioprocess monitoring – and what still stands in the way?
If “routine” means more commonly implemented than traditional laboratory-based controls such as HPLC, we still have work to do. The constantly evolving development landscape presents challenges, and while CMOs and CDMOs are critical partners, many have not yet made the level of investment in real-time spectroscopy equipment and expertise needed to drive widespread adoption.
That said, at-line spectroscopy has proven to be a practical bridge. We have successfully used at-line applications at both BMS-owned and CDMO sites to realize immediate benefits while evaluating the need for full real-time monitoring. As continuous processing continues to advance across the industry, real-time monitoring is likely to advance alongside it, as the two are inherently synergistic.
What impact could broader adoption of real-time spectroscopic monitoring have on batch release, quality control, and manufacturing efficiency?
Any process improvement effort begins with data, and the quality of that data is fundamental to making meaningful decisions. This is where real-time monitoring offers its greatest strength.
Broader adoption enables richer data sets and creates more opportunities to optimize both quality and efficiency. In addition, real-time spectroscopic monitoring can contribute to process greenness – an increasingly important consideration throughout development and manufacturing.
What have been the most significant advances in PAT and chemometric modeling over the past decade?
From my perspective, one of the most meaningful advances has been the effort instrument vendors have made to ensure their technologies are truly ready for use in regulated environments. Earlier in my career, novel analytical innovations were often introduced without sufficient consideration for practical implementation under GMP conditions.
Today, vendors are far more attentive to safety, data integrity, and regulatory expectations. In the API space, for example, we operate in electrically classified areas that require equipment meeting stringent ratings, along with appropriate materials of construction for direct process contact. Vendors have made significant progress in addressing these requirements, which has lowered barriers to implementation.
Are there emerging technologies – whether analytical, digital, or automation-focused – that you think will reshape pharmaceutical manufacturing in the next 5–10 years?
The most obvious development is the continued adoption of artificial intelligence. While modeling has long been part of real-time monitoring, the depth and breadth of AI‑based approaches are expanding rapidly. Advances in emerging AI tools may provide a deeper understanding of processes leading to better control, quality, and efficiency.
Automation is also becoming increasingly routine in both development and manufacturing. We have used automation to support activities such as method standard preparation and spectral data collection, reducing waste, time, and manual handling. At the same time, continuous manufacturing is emerging as a strong option for challenging processes that demand flexibility. These trends are likely to continue and reinforce one another.
Looking ahead, what does the future of biopharmaceutical manufacturing look like to you – and how central will analytical science be in realizing that vision?
The future of biopharmaceutical manufacturing is both complex and exciting. Our greatest challenge will be keeping pace with an environment that continues to evolve rapidly.
As complexity increases, so does the need for creative solutions that deliver meaningful process understanding. Advances in machine learning have lowered barriers to data analysis and modeling, but analytical science remains essential – it is the foundation for generating the high‑quality data those tools require. Expanding the use of real‑time spectroscopic monitoring will be critical to fully capitalizing on these advances.
After more than two decades implementing analytical technologies in production environments, what is the most important lesson you’ve learned about making new analytical tools succeed in industry?
It is never too early to consider how a spectroscopic monitoring solution might be implemented as an asset progresses through development. Solutions must be appropriate for the current stage while keeping sight of how monitoring will evolve over time.
Equally important is recognizing that the time and capital required must deliver a clear return on investment. While analytical scientists are naturally inclined to collect data, successful implementation depends on support well beyond analytical functions alone. A cohesive strategy, clear communication, and strong relationships with all stakeholders are essential – and the time invested in building those relationships pays dividends when challenges inevitably arise.
Newsletters
Receive the latest analytical science news, personalities, education, and career development – weekly to your inbox.
