How much of the process-related data generated by your pharma production provides real value? According to Marco Glauner, Head of Manufacturing & Supply Chain Solutions Europe at NNIT, a fresh look at existing manufacturing, laboratory and serialization data can reveal multiple opportunities for improving your business.
As members of a heavily regulated industry, companies in the life sciences sector usually excel at collecting and storing vast amounts of data as part of their manufacturing processes, especially when using software systems in those areas. However, utilizing that data in a productive manner is another matter entirely.
– Traditionally, compliance has been the main driver for collecting data in the life sciences sector. Pharma companies have gathered all sorts of process-related data in huge data lakes in case they need to provide evidence for an audit or product-quality issue. But in practice, finding relevant data in a vast, unstructured archive is often much more difficult and time-consuming than anticipated, says Marco Glauner, Head of Manufacturing & Supply Chain Solutions Europe at NNIT.
The consequence is that pharma companies end up spending significant resources on data gathering that barely fulfils the original purpose. But as Marco Glauner points out, the right combination of technology and industry insights can transform data management from a cost center to the foundation for data-driven decisions that provide substantial value to the business.
More relevance and better structure
Over the past decade, pharma companies have generally taken two approaches towards more productive data management. In the first approach, companies have focused on reducing the quantity of gathered data by assessing the relevance and impact on product quality. In the second, priority has been on applying more structure to the data in order to make it easier to utilize. The ideal scenario, Marco Glauner points out, is a combination of the two approaches.
– We encourage clients to take a proactive look at their data and use it for quality improvements, rather than only using the data when a problem occurs. Doing this requires both structure and impact analysis, which in turn require a deep understanding of pharma processes, Marco Glauner says and continues:
– It isn’t enough to be able to collect real-time data from a process and display it on a dashboard. In order to fully utilize your data, you must understand the impact of each parameter on the process or on other parameters. It may be acceptable if a particular value deviates by 2-3 percent, but a problem if the deviation is 10 percent. It also needs to be clear what initially caused such deviation – perhaps the deviation of another parameter? So, you must take a holistic view and analyze meta-data as well as relations to find the average values and know when to trigger alerts.
Build a solid data enablement strategy
Marco Glauner estimates that the majority of life sciences companies have the potential to benefit from a dedicated data-enablement strategy.
The first step is to map out your processes and ensure that you understand how each parameter influences them. Further, understand dependencies and relations between different parameters, even cross-system. This gives you a clear picture of which data you need to visualize on dashboards or feed into automated systems for AI-analytics.
Physical and chemical parameters such as temperature, pressure or pH value are the most obvious data-points to examine, but other aspects could also be relevant. For example, ensuring that all changes to the process are properly logged automatically to establish a reliable audit trail, then following any cross-dependencies – again, cross-system – that those changes might trigger.
Once you have identified your key data streams, you can select the best technology to accomplish your objectives. Perhaps you should use Microsoft BI to structure and visualize the data. Or it may be relevant to get real-time updates on batch production by connecting a Manufacturing Execution System (MES) or a Laboratory System (LES/LIMS) on top of a highly integrated production shop floor across the equipment.
– The main thing to remember is that the technology itself is not where the magic happens. The coupling between process understanding, knowledge of critical parameters and dependencies, system design and life sciences industry insights is the true enabler of data-driven and value-adding decisions, says Marco Glauner.
New uses for existing data
The opportunity for proactive use of existing data is not limited to compliance purposes. Marco Glauner points out that much of the data generated by pharma companies today can be used in new contexts to provide additional value. One example is the serialization data assigned to each individual unit of medicine during packaging.
– Pharma companies have complex systems in place to manage serialization data, because they must be able to account for the whereabouts of every single unit in case of a counterfeit suspicion. But that same information can be used for other purposes, such as tracking packaging efficiency, producing supply chain forecasts and planning logistics. Often, new opportunities can be identified by taking a fresh look at your data and breaking down internal silos to make data available to the right stakeholders within the organization, says Marco Glauner.
Finding the right business cases is essential, and it can be tempting to focus too much on the technology aspect. In many cases, a simple solution is often the most practical.
– Innovative technology will not make a difference if the business benefit is not the driving force. In the case of serialization, a simple barcode linked to a database provides all the necessary information. There have been several attempts to use RFID chips instead, but the problem was that they ended up being used as a more complicated and more expensive way to relay the same information as simple barcodes, says Marco Glauner and continues:
– The market missed out on the real value because they never properly implemented the unique benefits of RFID, such as providing real-time data on the status, the conditions (like temperature) and location of their shipments. The benefits of technical features were entirely missed.
