Digital transformation and Pharma 4.0 have become buzzwords in many industries, including pharmaceutical manufacturing. While the consensus is that these initiatives offer increased efficiency, cost savings, and better customer experiences, the realities of digital maturity are often overlooked and unplanned for, creating implementation delays and issues. Prioritizing and executing digital maturity correctly will result in advanced capabilities that generate value and lead to a sustained competitive advantage. In this article, we’ll look beyond the buzz, to the potential and realities of digital maturity.
The human side of digitalization
People are essential for digital transformation. The human element of Pharma 4.0 is so often overlooked in pursuit of digital excellence. This organizational transformation is just as important as the technology that you deploy, because the human process of change runs parallel to the process of manufacturing. It’s easy when focusing on Pharma 4.0 solutions to concentrate just on the technology and the digitalization and data pieces. To do so is to miss that, at its core, a lot of what we’re doing is still strengthening people, whether it’s a process engineer, a manufacturing operative, a systems analyst, or any member of the team. For this reason, the Applied Materials Pharma group focuses on developing intuitive, feature-rich user interfaces to make sure that the people using a solution easily access the data and functionality to enhance their productivity. Although the underlying technology can be really powerful, the user experience –the ability to use the software to navigate it, to see the data that someone needs in their exact role —is a big part of implementing successful digital transformation.
Digital maturity, digital transformation and data capture
Another often overlooked aspect of digitalization is the need to evolve the manual process. Often terms like digital maturity and digital transformation are used to describe simply going from an existing paper process to a digital version of it. That system needs to evolve and change to generate not just big data, but what’s been called ‘thick data,’ which has qualitative data points that can be used to help humans make better decisions, faster.
Once the process has changed and can create the type of data needed for decision making, the next challenge is to use it in a way that reduces decision latency. Data analytics is a cascade of information that can be a little overwhelming. Whether the manufacturer experiences a scheduling problem, a quality problem, or a maintenance issue, there’s often a social process around decision making. An engineer looks at the data around the event and others check it against process knowledge and other plant priorities before a decision is made. That can take a lot of time, or “decision latency.”
To reduce this with new product development, we first look at the process from end to end. Having that deep knowledge of the process means, by the time the customer is manufacturing, we’ve already built models that have been cleaned up and are far more user-friendly to aid decision making.
In the case of mature products, we look to see if we can use machine learning. This is because there are years of data already available that can potentially be used to iterate and improve within already validated processes. In this way, we gain value very quickly as the model learns and applies those lessons to continually optimize the outcomes.
Use case: scheduling
Scheduling can be challenging but is also ripe for innovation. As with decision making, scheduling is also often a very social process. There are daily meetings around a schedule which is still largely contained in Excel or a limited software suite. We’ve worked with our customers to integrate that data into sources such as their manufacturing execution system and training system so they can see what is happening in real time—what equipment needs to be ready, what person is operating it, etc. These kinds of use cases mean there are fewer meetings, which everyone loves, but more importantly that decisions are made much faster.
In these projects, we capture knowledge from senior engineers that really know the machines and the process and make it available within the system. This gives credibility to predictions and enables the less experienced engineers to be more effective because they have institutional knowledge to work with; they can access the “engineer in a box” experience. It’s taking a best practice from one place and moving it across the system.
Iterative process
Applied Materials has a long history working with the semiconductor industry, which has grown substantially over the last three to four decades. What we’ve been able to do is take what we learned from those partnerships and bring that to pharma along with a lot of the same technology for integrating data sources and automating decisions in real time.
Given the number of years it can take to bring a product to market, one of the things we are trying to do with our pharmaceutical clients when working on digital transformation is to really look at it as a long-term investment. Often, we want those six-month or one-year turnarounds on value. While sometimes these happen, looking at digitalizing processes, improving them, building better quality data sets, and assessing existing and new systems, is an iterative approach. Things may not go the way you expected, but every time you iterate, you’re making more data available. At the same time, people are growing and learning how to use these tool sets. This means we may not be looking at just a few months, but two or three years down the line. When we take that iterative approach, there are milestones and review points along the way to verify that we are driving toward the best outcomes for productivity and for the patients.
Scalability
Pharmaceutical manufacturers range from startups with a few people working in one space, to 100-year-old legacy companies with facilities on several continents. Obviously, each manufacturer along this spectrum will have a very different experience with digital transformation. Those who have a server in the same room as their experts are going to be agile, able to iterate quickly. Once they start to scale, though, it becomes crucial to have solid, documented processes, not only to add the value into their next product but to ensure traceability. There’s a fine line between wanting to be agile, but also being accountable. In these cases, it’s often worked well to have a group of people dedicated to innovation, to being on the shop floor learning and looking for ways to innovate in partnership with the people who run the processes every day. Out of these partnerships have come some innovative, effective and readily accepted solutions.
Larger organizations may have bigger challenges. They have IT systems, some of which go back decades. When advancing automation in these cases, it’s generally putting something new on top of legacy systems. In these cases, technologies like a service bus that extrapolates from the source systems and centralizes the data can standardize how the different systems communicate with one another so we can deploy new, powerful digitalization systems.
