Director of Digital Solutions
Endress+Hauser

Jason Pennington

At INTERPHEX 2026, Mel Radford, Bethany Silva, and Jason Pennington explore data trust, cybersecurity, organizational barriers, and KPI-driven thinking in smart pharma manufacturing adoption.

Why Loving Your Problem Is the Key to Smarter Pharma Process Optimization

Watch Part 1 of this interview on how smart sensors, connectivity, and advanced analytics are enabling predictive, insight-driven pharmaceutical manufacturing.

, Jason Pennington and Bethany Silva of Endress+Hauser and Mel Radford of Rockwell Automation discuss the barriers to smart instrumentation adoption and how manufacturers can measure meaningful progress in process optimization.

The panel opens by framing data trust as foundational. Manufacturers need data not only for reporting but to engineer new processes and procedures, and that data must be consistent, traceable, and audit-ready to satisfy regulators. The speakers emphasize that regulators are increasingly supportive of innovation and transparency, but companies must be able to clearly explain the "why" behind their data and decisions. Bringing offline or asynchronous analytical measurements into alignment with time-synchronized process data within a unified namespace is essential to achieving that standard.

On barriers to adoption, the trio identifies several. Financial constraints prevent some facilities from fully re-instrumenting and revalidating. Cybersecurity is a growing concern, as new field devices increasingly come with built-in IP addresses, requiring deliberate network segmentation and cyber resilience strategies as IT and OT environments converge. People and organizational priorities present perhaps the most persistent challenge, as internal quality teams can be more demanding than regulators and companies can struggle to prioritize among competing initiatives. The panel's advice is to engage quality stakeholders early, not at the end of a project.

For measuring success, the speakers point to throughput gains and manufacturing footprint reduction as the clearest KPIs. A real-world digital twin deployment across an entire drug product line serves as a compelling example, with one customer now rolling out the approach across all production lines after seeing measurable yield improvements. The panel closes with a reframe: rather than chasing solutions, start by deeply understanding the problem and defining how success will be measured.

Videos

At INTERPHEX 2026, Bethany Silva, Jason Pennington, and Mel Radford discuss how smart sensors, connectivity, and advanced analytics are enabling predictive, insight-driven pharma manufacturing.

Smart Instrumentation's Role in Pharma Manufacturing

, Jason Pennington, Director of Digital Solutions, Endress+Hauser; Bethany Silva, Industry Manager - Life Sciences, Endress+Hauser; and Mel Radford, Director - Life Sciences Global Accounts, Rockwell Automation, discuss with 

 how smart instrumentation, connectivity, and advanced analytics are transforming pharmaceutical manufacturing operations.

The panel emphasizes that while data from sensors is abundant, its value depends entirely on context. Simply collecting data is insufficient; manufacturers need a unified data model and a unified namespace that captures not just what an instrument measured, but when, how, and where, so that information can be meaningfully integrated across systems to drive real process insights.

Connectivity is identified as a foundational prerequisite, enabling data trapped in devices to flow into broader systems. However, the speakers stress that technology alone is not enough and that successful adoption requires change management and people-first thinking, since even the best technology fails without organizational buy-in.

On the instrumentation side, the conversation moves beyond traditional measurements like flow, pressure, and temperature toward advanced PAT tools such as Raman spectroscopy, which can provide molecular-level insight into cell growth and biological processes. This richer analytical data is essential for feeding and optimizing AI models and digital twins, reducing the number of design-of-experiment cycles needed in process development.

Perhaps most significantly, the panel describes a shift in manufacturing philosophy away from simple alarm thresholds toward predictive, trend-based process control. Rather than reacting when a parameter goes out of range, manufacturers can now anticipate deviations and intervene proactively, moving toward outcome-driven, policy-based process management.

Jason Pennington

Articles