- May 19, 2014
By Bill Lydon, Editor
A pharmaceutical company describes how they invested in multivariate tools and were able to gain better process understanding and improve batch outcomes. In effect, they were able to transform data to actionable information.
Perspectives from Pharmaceutical Automation Leaders – PAR Article Series, Part 4
By Bill Lydon, Editor
Pharmaceutical automation leaders from around the world gathered for the annual Pharmaceutical Automation Roundtable (PAR) in Copenhagen, Denmark, to discuss a number of automation challenges facing their companies. While the context was specific to the pharmaceutical industry, these challenges are certainly applicable to many industry segments. The topic of this article is multivariate analysis & control.
Biological processes are complex and difficult to predict. Some batches proceed normally and fail the endpoint critical product testing. The result is costly losses of material and time. The PAR presenter described how his company invested in multivariate tools and were able to gain better process understanding and improve batch outcomes. In effect, they were able to transform data to actionable information. These improvements required a combination of tools and people with the expertise to analyze the data and produce useful information.
The approach they deployed uses two main methods. One is offline gathering of data and subsequent analysis. The second is online real-time analysis.
The offline data is gathered from the ERP, LIMS (Laboratory Information Management System), process historian, and batch records. Batch records are still primarily paper and require manual data entry. However, they are in the process of being transitioned to electronic records. This offline information is powerful when using analytics for SPC, overlaying trends, and other analysis.
Online systems map out and trend the batch in real time using multivariant analysis. The data is plotted against pre-developed models. They have success building initial production models using development data from laboratory pilot scale runs. They normalize that data for production equipment. Even with clinical campaigns - when the product is made for the first time – the same process allows effective models to be developed for the first batch.
The tools are not only showing why batches fail within normal operating variables ranges. The tools provide further analysis and information. For example, if combinations of variables are running too high or too low, but are within normal operating variable ranges, batch failures are still occurring.
The tools also provide information about the interaction between process and equipment data that had not been apparent in the past. They monitor a range of variables, including pH, dissolved oxygen, temperature, pressure control valve positions, agitator speed, and temperature control valve percent open. When incorporated into a common model, this new data provides new insights. Everything is brought into one multivariant model. This helps find a range of problems, including maintenance issues.
The presenter provided an example of how this process enables fault detection and root cause diagnosis for troubleshooting. By starting with the high-level multivariate plot, they can zoom into individual plots to identify problems. In the example, a filter malfunction was identified that would not have been detected otherwise.
Organizationally, the company has an APC team with a steering committee sponsored by and chaired by the VP of Manufacturing. The team meets monthly with representatives from manufacturing, quality, automation, engineering, IT, technical development, and manufacturing sciences. Within the overall APC team, working teams are focused on offline, online, and technical investments. The cooperative analysis leads to site-based projects and initiatives.
There has been a big commitment to get the information into people’s hands. Large computer screens are set up in manufacturing areas to display the information. Online access allows monitoring anywhere. Staff can remotely monitor real-time multivariant analysis. In a real-world example, a problem was identified by a manufacturing manager from home who then contacted plant personnel to remedy it. The presenter emphasized that these methods have elevated everyone’s awareness level about what is going on during the process. Previously, the process seemed like a “black box,” and they had to wait for results at the end. They plan to implement alarms in the future to create more value.
The tools they use include standard off-the-shelf software. The offline tool is 100% supported by IT, and the automation interface provides the data sources. The online tool is a hybrid that fits between the IT and automation systems. The manufacturing sciences group supports it.
The company has defined and implemented a technology ladder with the goal to achieve parametric release of batches.
- Parametric Release
- Real-time Process Control
- Offline Process Control
- Multivariable Analysis
- Process Monitoring
- Analytical & Other New Technologies
The goal is more successful production batches. The presenter noted that a successful program requires management support and success stories.
These are comments from PAR participants:
We don’t have any instances that approach the online system described. We have had requests but the cost to setup these systems cannot be justified in our organization. Our people think the capabilities should be free with the tools they have. Historically, we did a lot of work with fermentation processes in the 1980s and 1990s. We developed a number of tools with some success and failure. There seems to be a lot of untapped potential value. Until we can get our engineering groups and technical services groups to the point where they can take advantage of these systems, we are not ready. Some areas are doing a good job with statistical process data.
We have two sites that have systems like those described. The tool is very functional and capable, but the tools are being analyzed again for their business value. There are some successes with these systems.
If your company has the right culture then these tools will work.
The move to multiproduct facilities will make this more valuable.
Our PAT (Process Analytical Technology) group is growing and they are implementing multivariable modeling and control. They have selected a vendor where they use centralized modeling software to minimize licensing costs. DCS companies are saying they can include the modeling software in their systems.
We are very good at looking at historic data when something has gone wrong. We are not mature with the online analytics to start implementing multivariant. It is an organizational issue where it cannot be justified at this point. It is something we should definitely be looking to use.
We have a lot of activity in this area. We are starting to roll out a new initiative to apply multivariate control with monitoring on CIP (cleaning-in-place) and SIP (sterilization-in-place). That is going very well. We have some ongoing pilots in R&D facility labs. It [control and monitoring] is viewed as an asset and we have some good success stories. In one case, we documented more than $1 million savings in a year by optimizing batches and achieving higher yields.
The scientists that really understand the process initially work with the multivariant tools. When they hand it over to new people, it helps them understand the biological interactions.
About the PAR Meetings and this Article Series
Every year, I have the opportunity to attend the Pharmaceutical Automation Roundtable (PAR) meetings, as the only outside observer. Last year’s meetings were held in September of 2013 at Novo Nordisk A/S at their facilities in Copenhagen, Denmark. Lead automation engineers from around the world attended this invitation-only, two-day event. This group of engineers has a wealth of practical knowledge and knowhow and is willing to share with other participants - truly learning from each other. The PAR meetings represent a very knowledgeable group of automation professionals gathered in one place at any one time to discuss automation issues. This year, the participating companies included Amgen, Biogen, Idec, J&J, Eli Lilly, NNE, Novartis, Novo Nordisk, Pfizer, Sanofi-Aventis. The PAR meetings consist of various presentations given by PAR members on specific automation topics. Other members then provide comments about their experience, ideas, and challenges relating to the topics. This article series presents a summary of those conversations with each article highlighting one or more of the topics covered by the PAR meetings. Comments by specific PAR members are reported anonymously.
PAR was founded about 15 years ago by Dave Adler and John Krenzke, both with Eli Lilly and Company. At the time, the purpose of the roundtable was to provide a means of benchmarking and sharing best practices for automation groups among peer pharmaceutical companies. The group specifically does not discuss confidential or proprietary information, cost or price of products, price or other terms of supply contracts, plans to do business or not do business with specific suppliers, contractors, or other companies.
Did you enjoy this great article?
Check out our free e-newsletters to read more great articles..Subscribe