Industry 4.0: The Invisible Hand
We all know that Industry 4.0 + Pharma = Pharma 4.0 – but are artificial minds and augmented realities changing anything tangible?
Angus Stewart | | Longer Read
Exactly 258 years have passed since the invention of the spinning jenny: the multi-thread device that replaced traditional sewing machines and paved the way for the first industrial revolution. Two and a half tempestuous centuries later, the fourth wave of industry is breaking upon our shores. But what can Industry 4.0’s cybernetic toolbox do for pharma in the 2020s and beyond? This is one of the big questions addressed by a new report from CRB Group (1), which breaks down predictions and preoccupations regarding the birth of “Pharma 4.0” and other emerging themes across 105 data-packed pages. To dive a little deeper into the next era of industrial technology, we spoke to CRB’s Senior Automation Engineer and Digitalization Lead, Yvonne Duckworth.
If there's an Industry 4.0, what were 1.0, 2.0, and 3.0?
All four stem from what we used to call the industrial revolution. The first industrial revolution saw the entry of machinery into production, and it was powered by coal, steam, and the birth of factories. The second industrial revolution was all about massively scaled-up production, powered by electricity and the assembly line factories most famously associated with the Ford Motor Company. The third industrial revolution was triggered by the addition of computers and mechanical automation – much of the work I do day-to-day has its roots in this phase.
Today, we’re living in the time of the fourth industrial revolution. This term – and the whole idea of there being four big transformative eras – began with a 2018 paper by Klaus Schwab, the founder of the World Economic Forum. His vision of Industry 4.0 is all about connectivity. It marks the arrival of networked digital technologies like smart manufacturing and cyber-physical systems.
Some of the devices driving these capabilities sit right at the cutting edge, while others came into being closer to the turn of the millenium. Under this rather wide umbrella we have robotics, AI, predictive analytics, RFID, the Internet of Things, cybersecurity, 3D printing, and pretty much anything using the cloud.
When a client tells me, “We want to incorporate Pharma 4.0 into our facility,” they could be referring to two of the above categories – or all of them. Pharma 4.0, just like Industry 4.0, is rhizomatic – the vision is one of a decentralized network more easily pinned down by an overarching theme than a neat hierarchical order.
Describe one strong example of Industry 4.0 in action…
One striking example sits just outside Greater Manchester, in the town of Macclesfield – not far down the road from The Medicine Maker’s UK headquarters, funnily enough. There, AstraZeneca runs what is known as a “lights out warehouse.” Put simply, that means it is fully automated to the point that people do not enter it. Everything within is dimly lit, and closely packed with very narrow aisles. This system leverages robotics to streamline handling, storing, and retrieving items in the warehouse, while interconnected to their enterprise resource planning system. To reduce risks of a fire hazard, the oxygen level is set to approximately 14 percent – much too low for human workers. Anyone who does venture inside must be equipped with the appropriate personal protective equipment (PPE) for a low oxygen environment. Perhaps my description of this robotic system sounds a little eerie from our human perspective, but it’s also a solid demonstration of what could be possible industrywide and what is being incorporated where the volume justifies the investment.
Where does pharma sit with Industry 4.0 more broadly?
Uptake has been slow. This is disappointing, but from another point of view it is hardly surprising as the pharmaceutical industry faces strict regulations compared with many other industries. In addition, implementing Pharma 4.0 is not simply a case of gathering all one’s desired technologies and switching them on. A more holistic approach is required – one which considers many different aspects including value, cost, and impact to workforce and culture.
First, one needs to understand each particular digital technology on a technical level. Then, one needs to understand what value each technology can offer from an ROI perspective, as well as evaluating what the associated costs are including initial investments as well as any annual recurring costs. Some technologies are more costly than others. Some technologies have varying degrees of impact to the workforce. In some cases such as automated warehouses, one can expect decreases to the workforce. Many can introduce new safety concerns. These are financial, ethical, and technical questions that no responsible company should rush past.
Data is another key piece to the puzzle. AI and integrated systems capable of communicating with each other can generate a huge volume of data. To analyze that data, it may be necessary to hire data scientists or data analysts. Recommendations from these experts may lead to a shift in the culture and structure of the company. Nothing is simple. Even the choice to go paperless has complex implications. Management has to be open to acknowledging these changes, as they must be the ones to implement them at every level – from top to bottom.
Is setup cost a serious hurdle?
As part of my job, I review the applicable technologies, evaluate which ones are best suited to recommend for CRB’s clients, and then help estimate the costs. These estimates can be pretty steep, depending on the technologies involved. Larger companies may be able to absorb such costs a little more easily, but the price tags are harder for smaller companies and startups. Often, they will choose to kick the can down the road, waiting for the day when their balance sheets can take the initial hit.
To players like this, I tend to recommend a phased approach. Rather than kicking that can down the road, a company can be smart and set up some of the necessary infrastructure in advance. Then, when more funds are available, everything is ready for the next steps in implementing Pharma 4.0.
Did the exodus into the internet during lockdown change the game?
During lockdown, we all became more virtual professionals, and that did make a difference. One example from my own work springs to mind; CRB had a client with a process equipment skid that was ready for a time-critical factory acceptance test. Under the “normal” circumstances of the past, our client would travel (often fly) to the on-site location of that equipment, and test it in person.
This time, our client was in a hurry. Travel restrictions had just come into place, so the usual fly-and-test routine wasn’t an option. It was at this point that they reached out to CRB for help. In response, we sent our client one of Microsoft’s HoloLens headsets, and worked with them to carry out a virtual factory acceptance test. Within 90 minutes of opening the box, our client was ready to begin the test. After that, we couldn’t keep up with demand! Every other client wanted to have a go at playing with the new normal.
How is the buzzword “metaverse” perceived within pharma, if at all?
The term “metaverse” isn’t really commonly used within the industry yet, but the way we are able to use virtual reality and augmented reality does work along those lines. The word captures the basic concept of working collaboratively in a virtual environment. Being able to see through the eyes of a colleague using augmented reality in another part of the world can be incredibly useful –not only for factory acceptance testing, but for remote troubleshooting as well.
At CRB, we use virtual reality when we are creating a 3D model design of a new facility. After donning a headset, you can see and explore everything you’ve designed. You can touch every panel, enter every room… and you need not do it alone. Clients anywhere can join you in experiencing a collaborative, walking tour of the facility, and share feedback by voice and chat. If you spot a fault while exploring your virtual model, you can draw a circle around it or add a note that everyone else will see.
How does AI sit within the Pharma 4.0 framework?
For a commonly-used, straightforward way to assess the progress of Digitalization in different companies within the industry, we can utilize a guide developed by the BioPhorum Group called the Digital Plant Maturity Model (2). This model has five levels:
Level One – The facility is based on predominantly manual processing with a low level of automation, as well as being primarily paper-based. Applications are standalone with minimal or no integration.
Level Two – The facility is based on “islands of automation” with digital silos. Some processes are manual, while some may be automated. Batch records may be semi-electronic or “paper on glass”. There is limited integration across functional silos.
Level Three – The facility is a more interconnected facility which includes vertical integration between enterprise systems such as ERP and MES along with the process control system layer. Data is collected in a site historian from process equipment as well as environmental monitoring. Electronic batch records along with review and release by exception are included. Industry standards such as ISA-88 (recipe) and ISA-95 are adopted.
Level Four – The facility is based on enterprise integration with internal integration from the plant to the value chain. Predictive facilities also include the integration of product development and manufacturing. This facility also includes proactive, predictive analytics which leverages AI to spot patterns and tendencies, model where they are likely to lead, and very often flag a warning if the predicted result could be dangerous or result in production downtime.
A Level Five facility is adaptive. This means it includes full end-to-end value chain integration from the suppliers to the patient. This is done in a modular, mobile and collaborative manufacturing environment. Level Five also includes a plug and play environment, zero downtime, and self-aware, continuously adaptive autonomous manufacturing. “Adaptive” also indicates the presence of AI software that is able to carry out analyses and then use them to adjust and self-regulate, further minimizing the need for human intervention.
Right now, there are no true fully adaptive and autonomous Level Five companies in existence. You do, however, see Level Five elements scattered here and there, which would lead me to believe that the rest won’t remain on the horizon forever.
Turning to another buzzword: blockchain. What can it do for pharma?
From my point of view and judging by the insights published in CRB’s report, data security could be crucial in the years and decades ahead because one of the great predicaments posed by Industry 4.0 is the proliferation of data. Running dozens of parallel devices and systems all in communication with each other, all producing highly accessible data, inevitably calls for a rebalancing via much more rigorous cybersecurity.
After all, there are horror stories. We have heard about pharmaceutical companies getting hacked and left with no option but to pay a hefty ransom to regain access to absolutely crucial data. Nobody wants to find themselves forced into that position.
At present, we aren’t seeing pharma make much use of blockchain, but I do think some of the larger companies are willing to pursue a certain level of implementation. For smaller companies, it’s not a popular option at this time. There are four main barriers stopping them: new security risks, implementation costs, computational costs, and validation.
Are there any other less headline-grabbing aspects of Industry 4.0 that also deserve a mention?
I think biometrics deserve more attention – and it comes with a much lower price tag, making it a great entry point for less titanic or wealthy organizations. A biometric device could be as small and simple as a smart watch, or even a bracelet. Such a device can grant the wearer the capability of logging into ten, twenty, or thirty different systems with a tap of the wrist. Coupled with this, that same wearable could also provide electronic signatures for process operations, as well as card access throughout a physical facility. That is just one very quick reason why I feel biometrics could do with a little more hype.
Web cameras need more love, too. Their veteran status in the toolkit can mean we easily take for granted the sheer range of applications these devices can have. While speaking to vendors, I have heard about some incredible ways they have been put to use. In the case of processing tanks, for example, setting up a networked camera inside the tank replaces the traditional task of lifting up the lid to check for foam buildup or CIP verification with remote monitoring. Cameras are also a great means for collecting data, as all the video streaming through the system can be archived and passed on to the company’s data historian.
I should also mention the use of 360° cameras as well as drones on construction sites. At CRB, we utilize 360° cameras to capture existing and current site conditions, as well as provide progress tracking. The 360° photos can be utilized for integration to the 3D design model using BIM 360. In addition, we have licensed drone pilots who are able to remotely monitor and scan construction sites with weekly flyovers for progress tracking of construction sites. Scans from these drones can then be translated over to the VR space, and fully integrated to the 3D design model.
Do you see any other major obstacles standing in the way of Pharma 4.0?
As you will see in our Horizons report, cost is obstacle number one. Risk management is another major concern. For example, issues to consider include risks to manufacturing operations, unscheduled downtime, safety, and security concerns to protect data and intellectual property. Trust in automated processes plays a significant role. For the implementation of Pharma 4.0 to move ahead, we will need to see trust in the safety and added value of the relevant hardware and software associated with digital technologies. Skillsets are another very valid concern. Operating new devices, managing new data, and carrying out validation on new systems could necessitate additional hiring and training.
As companies begin their journey to digitalization, there can be many barriers to entry. There is something of a scarcity of fully-tested technology available. Many companies are piloting new machines and systems, but the doors of the Industry 4.0 superstore remain partially closed – for the time being. That said, innovative technologies are continuing to rapidly advance the pharmaceutical manufacturing industry, breaking down many of these barriers. By following our roadmap and implementing a strategy for Pharma 4.0 adoption, companies can easily integrate digital technologies into the design of their facility.
Want to jump down a time tunnel and read what Yvonne Duckworth and a panel of experts had to say about Industry 4.0 in pharma, back when the term was newly coined? Check out our 2018 article, Towards Industry 4.0, and see for yourself where the ground has shifted.
- CRB Group, “Horizons Life Sciences Report 2021”, CRB (2021). Available at: https://bit.ly/CRB-Hor
- BioPhorum, “Digital plant maturity model (DPMM) version 2”, BioPhorum (2018). Available at: https://bit.ly/DPMM-guide