Context/Future Vision

Here, we consider how might the healthcare/pharmaceutical industrial ecosystem evolve in a predominant continuous manufacturing innovation and supply model in terms of changes to industry structure, adoption of enabling technologies, and the provision of new products and services to smaller patient group populations? Potential future developments within the medium term timeframe include:

  1. Patient (rather than health provider) centric supply chains that support multiple value and supply chain configurations, coexisting and providing different, often more localized and dynamic replenishment models.
  2. Simplified supply chain operations with less managerial oversight and regulatory sanction.
  3. More responsive production and distribution models that can support rapid replenishment driven by the emergence of patient management diagnostics and "Apps," medical devices, and supply chain integrating IT systems.
  4. Reduced capex and operating costs, and volume flexibility, afforded by continuous processing supports more geographically distributed production and supply networks, closer to patient demand.
  5. Process-control-based quality and regulatory assurance becoming established mechanisms, supported by advances in Process Analytical Technologies (PAT) that provide real-time data on product and process consistency during production, and used to quality assure product direct into the supply chain.
  6. Easier supply to smaller patient groups (by strength, by country) including earlier access to commercial scale materials for patients, as scale-up requirements become significantly less onerous. The reduced development timelines can increase the profitable supply time for innovators, allowing more development resources for the overall enterprise (assuming a supportive regulatory environment).
  7. More localization, enabling more dynamic closed loop control, with control parameters set upstream based on downstream measurements.

In terms of future scenarios, we might imagine the progressive emergence of cheap robotics and microprocessor control as well as advanced, but cheap, sensor systems supporting new models whereby complex molecules and bio-pharms are synthesized on small modular platforms, and numbered up to scale (as required). The decentralization of manufacturing that would occur would be unprecedented in the chemicals industry.

Another example of future implementation paths is the use of hybrid 3D printing systems to produce configurable flow reactors with sensing, actuation, reaction processing, purification, and so on. 3D printing is a process where objects can be fabricated layer by layer, or part by part, allowing computer design and easy customization of architectures. 3D printers come in several flavors, as well as high-end commercial systems and affordable, open source, user customizable devices. For instance, Cronin et al. have shown it is possible, using open-source 3D printers, to "hack" plastic laboratory ware. However, the development of this "hackware" is allowing the development of hybrid devices where the 3D printer is used not only to construct a test tube for a chemical reaction, but also deploy/pump the chemicals into the test tube for the reaction and also customize the test tube to allow certain reactions to happen in different ways. This could even be extended to biologics by printing bioreactors. The 3D printer acts in two ways. Firstly, it can be used to fabricate the plastic-ware, or "reactionware" (the "flow system" in which the chemistry is carried out), and secondly we can use the printer as a robot to move chemicals around to do chemical reactions. The 3D model can particularly support the deferment and late customization supply models within secondary processing, by allowing near-to-end market final processing and customization.

The potential realization of the "modular" chemical factory would require a new set of standards allowing modular interchange from a physical, chemical, electronic, and software point of view. The natural consequence of this could enable the retasking of the "factory" to produce new chemicals or drugs ondemand with zero extra capital cost or investment. This vision requires a radical new integration of chemical systems and synthetic methodologies developed within this new paradigm. The ultimate outcome would be the development of largely software-only manufacturing work-flows whereby the physical system could be reconfigured electronically. (This is similar to peptide or DNA synthesis today but DNA synthesis is an order of magnitude more reliable than peptide synthesis due to the combinatorial problem of conditions for coupling and deprotection as well as purification.)

Again the move to more niche products, possibly with local supply will be enabled by smaller, flexible manufacturing operations. The cost of build when comparing with batch for scale manufacturing plants is potentially reduced by > 30%, with physical footprints reduced by > 25%, cost of operations reduced by > 30%; these modest quantifications of potential benefits are based on successful continuous manufacturing examples already implemented.

At the personalized medicines level, personal "pill" fabrication may have a rather novel application for the consumer. We might imagine a patient that has complex medical issues requiring multiple drugs with multiple dose variations over time. Remembering to take the correct drugs at the correct time is an increasing problem. The local use of "pill-printers" that would simply combine pre-formulated version of the drugs together in either a liquid or solid form using a liquid or powder handling robot into a single dose. The "printer" would be programmed with the prescription of the patient and the drugs mixed together in a binder matrix and then formed into the pill. This could have obvious benefits for the patient in terms of adherence to treatment regimes, improving compliance especially for elderly patients with complex medical conditions.

Final delivery models to the patient could by-pass the current specialist distribution and pharmacy network with direct delivery models, already developed in prototype packaging equipment packing halls, able to serve patients directly with individually named product prescriptions, with multiple products filled on the same line with accuracy levels exceeding the manual checking undertaken within current "pharmacy" models.

Experience from other industries suggest this type of transformation to more flexible localized operations based on changing process technologies and more customized solutions is possible. Transformation examples with analogous comparisons with sector level transformations in other industries include: computer-assisted processing and control in aircraft, decentralization of the printing industry, and transport.


Computer-Assisted Processing and Control in Aircraft

The E2E continuous processing supply chain will be per se technically very complex, most likely more complex than the current batch model. However, clever use of computerized procedures will enable a better management of these procedures. An analogy with aircraft is that the Airbus A380 is more complex to fly compared with a 1950s prop plane. However, modern control systems can stabilize flight dynamics, navigate, and control the system much better, such that crew can reliably fly this plane and operate it more economically such that mass tourism is now possible, whereas in the 1950s, it was a very much a luxury. The inherent complexity in continuous processing will drive adoption of computer-assisted processing, and greater scrutiny in quality management. We can expect significant benefits to arise from this intense computerization. In chemical synthesis, the emphasis may be on liquid processes instead of solid/liquid dispersions that are easier to control and handle. The processes need to become simpler and more robust in hardware and the higher demand for control can be accommodated in software. This gives rise to a common, and highly compatible, hardware installation and greater impact on software/control. Once accomplished, technical transfers can be as simple as sending data.


Decentralization of the Printing Industry

The change in supply network structure in moving to continuous processing can be seen as a bifurcation point, as although technically more complex at a processing level compared with the batch paradigm, there is substantial opportunity for significant automation. A historic analogy illustrates this point. Printing in Gutenberg's time was a relatively simple process at a technical level. Even in the 1970s, it remained a fairly simple process albeit complicated by the addition of the mechanical features of mass-production. The printing process as such was still a simple shaping exercise of a stamp of some sort, inking it somehow, and transferring that ink to a sheet of paper. Today's laser printing is technically much more complex, but it is universally available, even when consumers do not have the slightest idea as to what is going on inside the printer itself.

The overall global consumption of paper has risen significantly since decentralized printing became available, despite the increased complexity of the technology. It is useful to note that the first laser printers in the 1970s were research institution type activities and yet are now a ubiquitous commodity.

In this example, complexity increases substantially in the initial phase requiring technical specialization with production capacity in centers of expertise. However, as the technology matures, production is progressively decentralized and requires smaller footprints at each production site. Decentralization will then drive supply chain topologies in different ways, as QA aspects are managed through in-field real-time data, or in the case where intervention is needed, the "cloud"-based operator electronically sends the necessary (software) upgrade. Today, the printing industry is highly decentralized, with printers now regarded as consumer goods. The flexibility of the print-on-demand decentralized model has outperformed the per se much faster high-speed printing technology of the traditional print-shop. A similar evolution, of decentralized production with software-based QA and upgrade interventions may also evolve within the continuous manufacturing model.


Transport

In commercial road transport, vehicles are able to take a variety of routes and select multiple stopping points. Vehicles themselves can have many operating states; traveling at speed, stationary with engine running, stationary with engine turned off, and so on, providing flexibility on delivery route to be taken and the ability to adjust operating costs throughout the journey. In an airplane however, one must operate at a minimum speed, and after take-off typically fly at a predetermined speed and route. Problems during the flight's trajectory must be corrected during motion with limited opportunities to "take-stock" and make unplanned stops. Both road and air transport modes offer opportunities and limitations, and selecting the correct mode of transport, or combinations of the two modes is determined by the context and flexibility required. There is no doubt that different transport requirements, in terms of distance to be traveled and number of drop-points will favor air or road transport in different ways, and a smart combination of both approaches will best support a diverse supply landscape.

These future supply scenarios highlight how batch and continuous operating models have evolved in other industries. These examples serve to demonstrate how decentralized and highly controlled process technologies have influenced the evolution of supply chain models in other industries and the mindset required to make such changes a reality.