Breaking down silos in pharma’s digitalisation drive
Justin Eames, senior manager, APM Solution Consulting, Aspen Technology Inc, explains why moving away from siloed structures is key to helping make pharma digitalisation a commercial reality.
Collaboration and communication can be stifled by siloed working practices that are prevalent in so many large pharmaceutical companies. Strict hierarchies mean data scientists have little interaction with maintenance and technical teams or C-Level decision-makers. As a result, many innovative ideas never make it to market.
Businesses are often reluctant to take on new technology evaluations which could achieve a more rapid time to value, anxious that they might affect validation status in their often complex and long-established validation processes. While Many large global pharmaceuticals businesses are innovative in new drug development, they are less innovative in manufacturing, relying on existing expertise and processes. Being less receptive to disruptive digital innovation means they can often miss out on efficiency-boosting initiatives and other benefits.
An industry tipping point?
The pressures of the Covid-19 pandemic could be a pivotal point for the industry. With existing manual processes are less easy to execute thanks to social distancing, now could be the time for the industry to shift its focus away from a conservatively academic outlook and on to applied innovation.
In the post-Covid world, it’s likely we’ll see more exploration of new digital initiatives to ensure plants can keep running should a second wave or something similar create further disruption. Some manufacturers are also recognising that they can operate efficiently with less people on site which could lead to streamlined operations.
Manufacturers are more likely to be receptive to advanced technological solutions that in no way directly affect the patient. So long as innovation does not impact drug manufacturing processes, but can enhance asset management, pharmaceutical companies are more likely to consider such solutions.
Supporting security of supply with predictive analytics
One of the biggest drivers for pharmaceutical manufacturers – especially in these turbulent times – is the security of supply.
Reducing supply chain disruption, reducing batch losses and increasing capacity of batch production are key areas of focus for many manufacturers while reducing CAPEX and lifecycle maintenance costs also come under intense focus. This means manufacturing equipment availability must be a top priority. Without exception, pharmaceuticals manufacturers tell us they want to be able to predict asset degradation and failure well in advance of an impending breakdown or other disruption, to be able to make decisions that can not only minimise cost and disruption but can also ensure continuity and resulting quality of drug supply, helping to protect public health.
The digital technologies behind this capability can also enable manufacturers to get more capacity from their existing equipment and prevent them from having to buy costly replacements, or even avoid large CAPEX investments building new manufacturing capacity.
Predictive capabilities are key to delivering real commercial benefits, providing the added value that pharmaceuticals manufacturers need over simple data analytics. This will likely help them to overcome the reluctance to take risks and persuade them to invest in driving greater efficiencies in the future.
Yielding greater results through innovation
Ease-of-use is one of the biggest opportunities of today’s predictive analytics and machine learning solutions. These enable pharmaceuticals companies to achieve rapid results without a single line of code having to be written. The data science is hidden and enables workers with little or no dedicated data science expertise to manage the solutions meaning existing staff can simply be trained to manage the platforms no re-staffing is necessary.
Managing the data itself is also relatively easy. The process of manufacturing drugs is hyper-controlled meaning there is little opportunity for ‘erratic data’ and the adoption of machine learning means fast results and real value can be achieved within weeks of implementation.
Multi-use models can also be advantageous in optimising similar equipment is in use across the pharmaceutical industry – such as a specific type of pump used in multiple services, or across several different production lines producing similar products. By sharing the normal and failure behaviours of assets found on a single machine with the other members of the pool. In doing so, save millions of pounds in value across the industry, rapidly increasing the scale and safety of an operation and avoid breakdown of all equipment of the same type and configuration.
Time for change
The predictive analytics capabilities outlined above are ideally suited to pharmaceutical companies and the restrictions put on manual working by the pandemic are a catalyst for change. This is the moment for pharmaceuticals manufacturers to make that necessary change.
Manufacturers need to avoid the cycle of conducting data science experiments that don’t positively impact their business. They should be starting to move their digital initiatives and experiments to real world commercial implementations that instead have a positive impact.
As the world continues to change at a rapid pace, businesses that choose not to adapt will be outpaced and outperformed by those that recognise and utilise the opportunities to drive commercial change, which have been made real by the latest digital technologies.