Distribution-independent inventory approach under multiple service level targets
A system and method are disclosed for an inventory planner that generates an inventory policy using any form of demand distributions, non-linear cost functions and/or multiple target measures of service levels, while taking into account a supply order lead time, such as, for example, a static or stochastic supply order lead time. The inventory policy generated by the inventory planner comprises an optimal and reproducible solution to one or more supply chain planning problems.
1. A system, comprising:
a computer comprising a processor and a memory, the computer configured to:
generate one or more transitional matrices comprising a supply leadtime;
generate one or more reward matrices comprising cost components;
model a supply chain planning problem as a Markov decision process (MDP) based, at least in part, on the generated one or more transitional matrices and the generated one or more reward matrices;
solve the MDP, by:
determining, from solution values where an order quantity is zero, a resulting state after an order action is applied;
generating from the solution values, indices comprising a minimal state and a target state;
generate an inventory policy based, at least in part, on the solution of the modeled supply chain planning problem wherein the computer adjusts inventory levels of at least one stocking point and adjusts sourcing of one or more supply chain entities, based at least in part, on the inventory policy; and
in response to and based, at least in part, on the inventory policy, cause the one or more supply chain entities to ship one or more items.
2. The system of claim 1 , wherein the computer generates the one or more transitional matrices by:
generating a set of inventory states comprising a maximum possible demand and a maximum inventory quantity; and
setting a probability of staying in state s equal to a probability of not having a demand, and setting a probability of moving from state s equal to a probability of having d demand.
3. The system of claim 2 , wherein the computer is further configured to:
when the supply chain planning problem is modeled according to a time dependent model, parameterize the one or more transitional matrices by a probability associated with the supply leadtime; and
when the supply chain planning problem is modeled according to a time independent model, aggregating the one or more transitional matrices according to the probability associated with the supply leadtime.
4. The system of claim 3 , wherein the computer generates the one or more reward matrices by:
associating a cost with each inventory state i for each ordering action q at a time step t.
5. The system of claim 4 , wherein the cost comprises one or more of inventory cost, order cost, and stockout cost.
6. The system of claim 5 , wherein a target service level is modeled as a penalty and added to the cost of the one or more reward matrices.
7. A method, comprising:
generating, by a computer comprising a processor and a memory, one or more transitional matrices comprising a supply leadtime;
generating, by the computer, one or more reward matrices comprising cost components;
modeling, by the computer, a supply chain planning problem as a Markov decision process (MDP) based, least in part, on the generated one or more transitional matrices and the generated one or more reward matrices;
solving the MDP, by:
determining, by the computer, from solution values where an order quantity is zero, a resulting state after an order action is applied;
generating, by the computer, from the solution values, indices comprising a minimal state and a target state;
generating, by the computer, an inventory policy based at least in part, on the solution of the modeled supply chain planning problem wherein the computer adjusts inventory levels of at least one stocking point, based at least in part, on the inventory policy and adjusts sourcing of one or more supply chain entities; and
in response to and based, at least in part, on the inventory policy, cause, by the computer, one or more items to be shipped from an inventory.
8. The method of claim 7 , wherein generating the one or more transitional matrices comprises:
generating, by the computer, a set of inventory states comprising a maximum possible demand and a maximum inventory quantity; and
setting, by the computer, a probability of staying in state s equal to a probability of not having a demand, and setting a probability of moving from state s equal to a probability of having d demand.
9. The method of claim 8 , wherein the method further comprises:
when the supply chain planning problem is modeled according to a time dependent model, parameterize, by the computer, the one or more transition matrices by a probability associated with the supply leadtime; and
when the supply chain planning problem is modeled according to a time independent model, aggregating, by the computer, the one or more transition matrices according to the probability associated with the supply leadtime.
10. The method of claim 9 , wherein generating the one or more reward matrices comprises:
associating, by the computer, a cost with each inventory state i for each ordering action q at a time step t.
11. The method of claim 10 , wherein the cost comprises one or more of inventory cost, order cost, and stockout cost.
12. The method of claim 11 , further comprising:
modeling, by the computer, a target service level as a penalty; and
adding, by the computer, the penalty to the cost of the one or more reward matrices.
13. A non-transitory computer-readable medium embodied with software, the software when executed is configured to:
generate one or more transitional matrices comprising a supply leadtime;
generate one or more reward matrices comprising cost components;
model a supply chain planning problem as a Markov decision process (MDP) based, at least in part, on the generated one or more transitional matrices and the generated one or more reward matrices;
solve the MDP, by:
determining, from solution values where an order quantity is zero, a resulting state after an order action is applied;
generating from the solution values, indices comprising a minimal state and a target state;
generate an inventory policy based at least in part, on the solution of the modeled supply chain planning problem wherein a computer adjusts inventory levels of at least one stocking point, based at least in part, on the inventory policy and adjusts sourcing of one or more supply chain entities; and
in response to and based, at least in part, on the inventory policy, cause one or more items to be shipped from an inventory.
14. The non-transitory computer-readable medium of claim 13 , wherein the software generates the one or more transitional matrices by:
generating a set of inventory states comprising a maximum possible demand and a maximum inventory quantity; and
setting a probability of staying in state s equal to a probability of not having a demand, and setting a probability of moving from state s equal to a probability of having d demand.
15. The non-transitory computer-readable medium of claim 14 , wherein the software is further configured to:
when the supply chain planning problem is modeled according to a time dependent model, parameterize the one or more transition matrices by a probability associated with the supply leadtime; and
when the supply chain planning problem is modeled according to a time independent model, aggregate the one or more transition matrices according to the probability associated with the supply leadtime.
16. The non-transitory computer-readable medium of claim 15 , wherein the software generates the one or more reward matrices by:
associating a cost with each inventory state i for each ordering action q at a time step t.
17. The non-transitory computer-readable medium of claim 16 , wherein the cost comprises one or more of inventory cost, order cost, and stockout cost.
18. The non-transitory computer-readable medium of claim 17 , wherein the software is further configured to:
model a target service level as a penalty; and
add the penalty to the cost of the one or more reward matrices.