Predicting Commodity Flow Shifts in Global Supply Chains under Changing Arctic Conditions, 2022-2065

With climate change has come a thawing of Arctic sea ice, creating a growing possibility of increasingly passable Arctic waters for longer periods of each year. This change creates opportunity for logistical alternatives in support of global supply chains (SCs), and potential for realignment of raw material and middle-product suppliers in manufacturing. This paper presents a methodology for predicting how SC might restructure with the opening of the Arctic passageways.
A stacked machine learning method that represents the multi-echelon, multi-commodity SC with only publicly available data is applied in predicting future trade flow changes.
The method was applied to predict changes in the SC for lithium-ion battery production under predicted future sailing conditions in the Arctic waters. Results of the case study suggest that Arctic routes could drive significant changes in trade patterns between nations, and therefore changes in SC structure. These changes could encourage countries like Germany to reduce reliance on certain exporters (e.g., China) in favor of closer or more cost-effective trading partners, such as Korea, Japan, and Poland.
This paper applies machine learning techniques to predict how changing Arctic conditions might impact global SC structure in the future and illustrates the methodology’s capabilities for an import product: lithium-ion batteries used in electric vehicles.