receive a current state of one or more items in a supply chain network by one or more imaging devices that generate a mapping of the one or more items in the supply chain network;

check for prior identification of one or more complicating nodes;

partition the supply chain network based on the one or more complicating nodes;

determine whether the partitioned supply chain network generates balanced subproblems;

in response to determining that the partitioned supply chain network generates balanced subproblems, identify a variable partition;

in response to determining that the partitioned supply chain network cannot generate balanced subproblems, partition the supply chain network based on determining that the supply chain network is convergent or divergent, wherein the supply chain network is partitioned as a generic supply chain network in response to determining that the supply chain network cannot be determined to be convergent or divergent;

partition variables of the balanced subproblems by removing variables from each of the balanced subproblems;

partition constraints by generating a list of constraints of the supply chain planning problem for each of the balanced subproblems;

formulate linear programming subproblems based on the partitioned variables and the partitioned constraints;

perform masterless iteration with subgradient descent to iteratively optimize the linear programming subproblems until one or more stopping criteria are reached to generate a solution for the linear programming problem; and

instruct automated machinery to adjust inventory of the one or more items based at least in part on the solution to the linear programming problem.