AI for Managing Vaccine Supply Chain

This blog post is part of the series on AI and vaccination. Also see:

How AI Can Help with Vaccination

Who Gets the Vaccine First? AI Helps but Humans Decide

Covid Vaccine needs Blockchain Transparency; U.K. Sees the Light

Supply chain is at the core of three gargantuan vaccination tasks – manufacturing, distribution and administration. The common goal is to vaccinate maximum people, at a maximum speed, and at a planetary scale.  How about that?

Gartner’s recent Future of Supply Chain survey identified artificial intelligence and machine learning as the top important and disruptive technologies.

Clarity about How AI Can Help with Vaccination surfaces the AI’s critical capabilities for supply chain. Who Gets the Vaccine First underpins the current unprecedented dynamics of the supply chain challenges and frequent turns. Some AI capabilities for supply chain are already common, well-understood and tested; this includes planning, sourcing, manufacturing, fulfilment, logistics and cross-functional collaboration. The question here is how to quickly implement the known AI approaches that deliver supply chain agility to save resources and maximize the speed of vaccination.  Examples of existing solutions useful for solving vaccination challenges include:

• Supply chain network mapping and visualization (for traceability, sustainability and transparency)
• Holistic inventory visibility (for better delivery accuracy, resource savings, capability improvement)
• Forecasting of all kinds (to improve forecasting accuracy, reduce inventory and storage costs)
• Anomalous order detection (to maintain and improve supply chain velocity and agility)
• Order probability and date prediction (to reduce inventory, shortages and backlog)
• Customer-level demand forecasting (for better alignment with customers in collaborative planning, forecasting and replenishment; reduced manual intervention)
• Material quality prediction (for cost and risk reduction)
• Contract review and analysis (to increase contract strength and velocity)
• Improved supply chain team member expertise (for higher productivity)

But many necessary AI capabilities are new, due to the conditions introduced by the pandemics and because of the mere scale of tasks. Such capabilities are delivered by artisan data scientists who develop custom models.  I know of at least one state that established the role of the Chief Data Scientist because of the impactful work and insights delivered with AI and machine learning. Examples of newly developed AI solutions for solving vaccination challenges enable participants to:

• Create, manage, and track care plans for vaccination phases
• Determine state/regional/ location/site-based vaccine needs based on COVID spread and waves
• Dynamically manage supply chain for drugs, hospital equipment, and non-drug materials depending on the state of vaccination
• Predict, manage and balance supply of non-vaccination products, such as oxygen
• Proactively manage point of use storage and supplemental supplies (syringes, alcohol, vials)
• Monitor temperature controls and shelf life of the vaccine

As a big proponent of graph technologies, I would also like to point out the effectiveness of graph technologies in the supply chain logistics, full visibility into a pharmaceutical manufacturing process, and PPE planning and allocation in the healthcare facilities. No wonder, we saw a massive rise of graph machine learning this year.

I want to express my deep respect and gratitude to those data scientists and many-many other specialists who work around a clock, without weekends, often sleeping 5-6 hours a day. We have very few public examples, since all participants are busy delivering solutions and capabilities, rather than working on reference nuances and legalities. I am sure we will learn much more later, so forgive me if I shortened and anonymized the examples. All I can say, I saw them all.

Gartner Coronavirus (COVID-19) Resource Center

Follow Svetlana on Twitter @Sve_Sic