By Anna Codina
The current global supply chain issues are clearly an unresolved challenge that the pharmaceutical industry is also facing. The pandemic has put under stress supply chain and manufacturing forcing the increase of agility, mobility and adoption of new technologies regardless of the regulatory burden that this implies. The concept of portable, continuous, miniature and modular manufacturing is a reality, [1,2] but perhaps not yet accessible to the pharmaceutical and biopharmaceutical ecosystem.
Bruker has therefore joined forces with 3 other companies (NovAliX, Alysophil, De Dietrich Process Systems) to take on-demand API manufacturing a step further.[3] The partnership leverages the combination of breakthrough synthesis, continuous flow chemistry, in-flow benchtop NMR and IR analysis with artificial intelligence and additive manufacturing to create an autonomously AI-piloted chemical micro-plant. Benefits of this approach are twofold. First, the provision of on-premises molecule production, reducing the dependency of global supply chains and second, the improvement of safety in production by using small quantities of intermediate chemicals, reducing the danger zone in case of accidental release.
This talk gives an update on achieved milestones towards the progression of the project, including demonstration of the first real-life synthesis control powered by deep reinforcement learning. The AI agent’s ability to adjust the reaction conditions on-the-fly is particularly useful for reactions that are difficult to control or that require precise conditions to be successful.[4] The key role of benchtop NMR, as integrated PAT tool, for process optimisation and control will be demonstrated.
1. Miyai Y et al., Org Process Res Dev., 25 (12), 2707-2717 (2021)
2. https://www.pfizer.com/sites/default/files/investors/financial_reports/annual_reports/2019/our-bold-moves/deliver-first-in-class-science/portable-continuous-miniature-modular/index.html, accessed 10Dec22
3. https://www.dedietrich.com/en/news/corporate/pipac-collaboration-enables-new-compact-mobile-and-frugal-approach-active, accessed 10Dec22
4. https://ala.associates/projects/alysophil-demonstrates-first-real-world-solution-in-ai-powered-autonomous-chemical-synthesis-for-apis-within-pipac-project/, accessed 15Feb23