Hierarchy of Pharma MES needs
Pharma or BioPharma drug Manufacturing is a process that needs accurate measurements, precise actions in exact sequence while recording it for audits or people die. For years I’ve thought how a Pharma MES can unleash the potential in this space. MES is proven in many parts of the world and yet, there’s widespread use of paper for Instructions and records. Why aren’t manufacturers lining up for getting one?
Part of the story (and more so in developing world) is Cost. However with the risk of 483s, costs can be more justifiable. There can also be reasons of limiting data exposure to minimum required, because an MES would increase transparency beyond comfort. Lastly, almost all the time I’ve talked to a manufacturing floor person, they’re always convinced about MES but it has to be tailored to their needs.
That’s not new, isn’t it? MES 101 says — Its very industry specific
To cover all industries, MESA might define the role of MES to include all possible functions there can be. However, if we sell all the bloatware that has scheduling to labour management, we’re doing it wrong. Even if we go with the MESA’s earlier model with 11 major functions, each functions matters differently to each industry or even plant. Simply because:
- If it ain’t broke, don’t fix it. If you had another system taking care of a function of MES and was doing it well, you may not want to yank that out unless you’re getting huge benefits. E.g. A lot of the customers might already be using specialized scheduling software for their batches and thus might not want to make a switch
- Need in some areas might just be enhanced because of regulations. e.g. Quality traceability is an important aspect in Pharma.
- Lastly, not every function provides the same bang for the buck. Say maintenance is a low-key activity for you, implementing an integrated system within MES might be an overkill
So lets say, by making it industry focused, we reduced the MESA model to a more manageable number. Still are all the remaining functions equally important?
This is where I come to what I call The Maslow’s hierarchy of Pharma manufacturing needs
What Maslow established long back in Psychology, applies to the pharma manufacturing. Even though this model could be applicable to each site and country level, but largely, the bigger the pharma company, the higher up they are in the hierarchy.
Hierarchy starts with basics of getting manufacturing going and producing drugs profitably
Data Acquisition and Integration
Any manufacturing starts off with a bunch of equipment that need to be controlled. This portion primarily deals with the interaction of MES with different layers in Automation pyramid. MES needs to collect data from DCS or SCADA systems to know automatically where possible. Interacting with ERP to get work orders and report inventory utilization and interacting with a plethora of systems like LIMS, Weight scales, RFID scanners, Integrity testers. The list of interfaces is ever increasing, and more integration ensures that we’re able to increase the perimeter of events which MES is aware of. What do we do with this newfound information? This takes us to the next point
Compliance and Data Integrity
This is the main reason pharma industry entertains an MES expense in the first place. Now that we know raw data about events on the shop floor, we can control better and provide better guarantees of repeatability, robustness, and quality of the drug. Compliance and quality buckets cover the aspects of Genealogy, tracking deviations, reviewing, reporting, and releasing. An electronic closed loop system ensures that no activity happens outside the perimeter that wasn’t captured in quality purview.
Electronic Work Instructions
One important contributor to Biopharma manufacturing is: us, humans, and we’re not covered under the machine to machine interfaces (well at least till neuralink has figured wiring us to a computer). Thus, Electronic work instructions or Operator guidance provides a Human machine interface that reduces transcription errors, production mistakes and even builds enough interlocks to ensure that the man and the machine work in harmony and safety.
So the big caveat now
- I am not at all saying that anything on the top of this hierarchy isn’t important. They absolutely are. Some will even start at a higher level because of their unique circumstances but the majority might follow this course to self actualization.
- There are a lot of overlaps within the categories. E.g. Audit trail is an important part of Electronic work instructions and also an integral part of compliance and quality.
- There are 3 different Users of MES (Sorry! Personas)and so depending on which department is pushing digitization, you’d see that need moving to the base of the hierarchy. e.g. Interfaces are typically the IT or Engineering team forte, Compliance and Quality functions within MES address the quality group needs and the Electronic work instructions serve best the Manufacturing organization and the Technical services.
While Maslow was on my mind, I also stumbled over an excellent thread by Shreyas Doshi on Twitter where he talks about the Focusing Bias
Nothing summarizes it better than the thread itself. Like each Manufacturer is at a different level on the Hierarchy and thereby values each of the MES functions differently, it can be represented with a Customer Problem Stack Rack (CPSR). MES functions can be stacked in terms of the problem it solves for the single plant.
It quite resonated with me. You could feel that after each Voice of Customer (VoC) exercise, there’s no mismatch between the MES product offering and what customers want and yet not see a huge uptake. For example, everyone agrees they want to build a golden batch with continuous feedback to improve yields. No customer ever downplayed OEE. However as quoted in the thread,
“Nothing in life is as important as you think it is, while you are thinking about it” — Danny Kahneman
i.e. Anything customers told you as important, is only one of their problem and not their top problem for which they are looking for solutions in the market.
Now getting each customer a different product addressing their different Problem Stack is a dream but it should be approachable. How we cater this diversity of needs is a matter of a different post (hopefully soon as I’ve procrastinated writing this for quite a long time)
For now I think segmenting the market such that you see different manufacturers at different levels of hierarchy and unbundling your product to serve individual needs can be good approaches. I am sure there’s a lot more to it. Please feel free to leave comments on your thoughts and I’d be happy to collaborate on building more such models.
-Apurva
PS: All these are personal views and opinions only
