Subscribe to Blog Updates

Shifting from systems to a service-oriented architecture

When the World Health Organization (WHO) declared COVID-19 as a pandemic in March 2020, the pharmaceutical sector burst into action. By May, there were over 1,000 clinical trials ongoing and by August, 167 vaccine projects were underway. By nature, the pharmaceutical industry is driven by development. Yet, pharmaceutical engineering has not always kept pace. Here, my colleagues from ISPE Pharma 4.0 – Plug & Produce working group Anton Granget, Kim Hewson, Martin Kuckelinsky and I would like to explain why digital integration in pharmaceutical manufacturing must evolve.

    

Despite its advances in medicine, pharmaceutical’s plant architecture has remained relatively unchanged in the last few decades. These sites typically operate with legacy equipment and often use traditional automation architectures.

This architecture uses separate levels of automation for different types of system. For instance, business applications like enterprise resource planning (ERP) and product lifecycle management (PLM) will operate in one silo, while MES or processing monitoring applications, like SCADA software, will remain in another.

However, as connectivity and integration become ubiquitous in other manufacturing environments, there is an opportunity for pharmaceutical to follow suit. So, why, in an industry so heavily associated with advancement, is there hesitation?

Traditional architectures

With its complex regulatory requirements and heavy reputational consequences should something go wrong, the pharmaceutical industry is certainly more risk-adverse than other sectors.

While traditional automation architecture may be tried-and-tested, it does not provide an optimal way to use data in modern facilities. Let’s say a plant operative wants to remove inefficiencies from the manufacturing process. To determine areas for improvement, the operative would need to extract subsets of data from different systems, such as the manufacturing execution system (MES), Laboratory Instruments Management System (LIMS), ERP and SCADA.

In this example, an operator would need to take multiple steps between laboratories, production and enterprise space to gather all of this information – that’s not to mention the difficulties they may face when attempting to collect data from unfamiliar systems.

In an ideal world this data would be integrated into a harmonized, easy-to-understand interface, providing the operative with a ready-made method to spot inefficiencies. There are already technologies that enable some integration for pharmaceutical manufacturers, but this software integration by dedicated interfaces tends to come at a high cost and doesn’t provide the full data harmonization required to meet today’s needs.

In truth, the industry needs a complete shift from this architecture type and a move to a services-oriented structure where integration is the rule, not the exception.

Service-oriented structures

A service-oriented architecture will operate as a grid of modular micro services. As opposed to a collection of hard-wired systems, this architecture model will allow more flexibility and agility by giving users easy access to a wider pool of data.

Consider it like making a meal in a kitchen. While you may know exactly where to find your ingredients and equipment, the process of collecting these items from separate areas of the kitchen is far more longwinded than having access to a pre-laid buffet. That’s not to mention the painstaking task of cooking — or in this case, analyzing the data — when you’ve retrieved your items.

By choosing a buffet of data, as enabled by the service-oriented approach, pharmaceutical manufacturers can reap considerable benefits. The model allows for more flexibility, by enabling the quick assembly and disassembly of systems in the architecture. It also enables better scalability than previous models by ensuring that the data can be reorganized and presented in a way that’s beneficial to whoever is looking at it.

Returning to our earlier example, an operator will no longer need to investigate what system each dataset resides on. Instead, the data can be distributed in various services and accessed from a single interface, regardless of where it is located.

Most importantly for the pharmaceutical industry though, this architecture model can also be achieved while maintaining necessary cybersecurity and Data Integrity compliance requirements by automating configuration and validation processes.

Time to switch pace

At the time of writing this article, more than 170 teams of researchers are in the process of developing a COVID-19 vaccine — twelve of which are already in phase three efficacy trials. As the pandemic has demonstrated, the pharmaceutical industry moves fast, and the technologies used in pharmaceutical manufacturing should be advancing too. The ISPE Pharma 4.0 initiative is working to ensure this happens. The group is designing new architectures to accompany the pharmaceutical sector towards more flexible, interoperable and sustainable production. Pharmaceutical manufacturers must embrace a paradigm shift in how automation architectures are traditionally constructed to keep pace with the industry’s product development, moving away from a system-oriented world into a services-oriented architecture.

 

The four authors of this article are members of the ISPE Pharma 4.0 – Plug & Produce working group.

Anton Granget – IIoT consultant – TechPivot GmbH
Kim Hewson – Automation Solution Director – GlaxoSmithKline
Martin Kuckelinsky – Head of Automation Engineering – Takeda
Giuseppe Menin – Industry Manager Pharmaceutical – COPA-DATA