Decision support system for optimizing the unit identifier stocking decision
Technology for optimizing automated inventory management is described. In an example implementation, a method, implemented using one or more computing devices, may generate unit identifier sets based on affinities between items uniquely identified by the unit identifiers, determine first and second contribution values for each of the unit identifier sets, and calculate a difference between the first and the second contribution values for each of the unit identifier sets. The method may further calculate an incremental carton and expense prevention opportunity (ICEPO) value for the unit identifier sets, calculate an adjusted fulfillment center stocking score using the ICEPO value and the difference between the first and the second contribution values for each of the unit identifier sets, select a subset of the unit identifier sets based on the adjusted fulfillment center stocking score, and output the subset to an inventory replenishment system that manages stocking of items.
1. A computer-implemented method comprising:
generating, by one or more processors using order data describing items uniquely identified by unit identifiers, one or more sets of the unit identifiers based on one or more affinities between the items;
calculating, by the one or more processors, an incremental carton and expense prevention opportunity (ICEPO) value for each of the one or more sets of unit identifiers based on a comparison of fulfillment of a set of unit identifiers from a reference fulfillment center using a single carton with fulfillment of a first unit identifier of the set of unit identifiers from the reference fulfillment center using a first carton and fulfillment of a second unit identifier of the set of unit identifiers from a second fulfillment center using a second carton;
calculating, by the one or more processors, an adjusted fulfillment center stocking score using the ICEPO value and a difference between a first contribution value and a second contribution value for each of the one or more sets of unit identifiers;
selecting, by the one or more processors, a subset of the one or more sets of unit identifiers based on the adjusted fulfillment center stocking score;
computing, by the one or more processors, forecast data using a machine learning model trained using the order data;
generating, by the one or more processors, an electronic data signal based on the selected subset of the one or more sets of unit identifiers and the computed forecast data; and
performing one or more of providing a graphical display using the electronic data signal and transmitting, by the one or more processors, the electronic data signal representing the subset of the one or more sets of unit identifiers to an inventory replenishment computing system via an application programming interface, the inventory replenishment computing system managing stocking of items uniquely identified by unit identifiers.
2. The computer-implemented method of claim 1 , further comprising determining, by the one or more processors, the one or more affinities between the items uniquely identified by the unit identifiers based on the order data.
3. The computer-implemented method of claim 2 , wherein determining the one or more affinities between the items uniquely identified by the unit identifiers based on the order data includes
determining, by the one or more processors, a threshold frequency for identifying sets of unit identifiers,
determining, by the one or more processors, groups of unit identifiers using the order data, each group of unit identifiers appearing within at least one individual order, and
determining, by the one or more processors, a quantity of individual orders in which each group of unit identifiers appears.
4. The computer-implemented method of claim 3 , wherein generating the one or more sets of unit identifiers based on the one or more affinities between the items uniquely identified by the unit identifiers includes
selecting, by the one or more processors, the one or more sets of unit identifiers from among the groups of unit identifiers based on the quantity of individual orders in which each group of unit identifiers appears and the threshold frequency for identifying sets of unit identifiers.
5. The computer-implemented method of claim 1 , wherein transmitting the electronic data signal to the inventory replenishment computing system includes communicating the electronic data signal to the inventory replenishment computing system via a computer interface, the inventory replenishment computing system configured to stock items in the reference fulfillment center based on the electronic data signal.
6. The computer-implemented method of claim 5 , wherein the inventory replenishment computing system is further configured to stock items in the reference fulfillment center based on a forecasted demand of the items for a geographic region served by the reference fulfillment center.
7. The computer-implemented method of claim 1 , wherein calculating the ICEPO value for the one or more sets of unit identifiers includes determining a delivery expense from a fulfillment center to a defined location for the items uniquely identified by the one or more sets of unit identifiers.
8. The computer-implemented method of claim 7 , wherein calculating the ICEPO value for the one or more sets of unit identifiers includes determining a delivery expense from an alternate fulfillment center to the defined location for the items uniquely identified by the one or more sets of unit identifiers.
9. The computer-implemented method of claim 1 , wherein selecting the subset of the one or more sets of unit identifiers based on the adjusted fulfillment center stocking score includes
sorting the one or more sets of unit identifiers into a hierarchy based on the adjusted fulfillment center stocking score, and
selecting the subset of the one or more sets of unit identifiers based on a threshold quantity of sets of unit identifiers, the hierarchy, and an excess stocking capacity of a fulfillment center.
10. The computer-implemented method of claim 1 , further comprising:
determining a relevance of the one or more sets of unit identifiers to current orders using the order data;
setting a frequency based on the relevance; and
performing one or more operations at the set frequency.
11. The computer-implemented method of claim 1 , further comprising:
determining, by the one or more processors, the first contribution value for each of the one or more sets of unit identifiers based on fulfillment of one or more units of the set of unit identifiers using inventory stocked in the reference fulfillment center;
determining, by the one or more processors, the second contribution value for each of the one or more sets of unit identifiers based on fulfillment of one or more units of the set of unit identifiers using the second fulfillment center; and
calculating, by the one or more processors, the difference between the first contribution value and the second contribution value for each of the one or more sets of unit identifiers.
12. A system comprising:
one or more processors; and
a non-transitory computer readable medium storing instructions that, when executed by the one or more processors, are configured to perform operations including:
generating, by the one or more processors using order data describing items uniquely identified by unit identifiers, one or more sets of the unit identifiers based on one or more affinities between the items;
calculating, by the one or more processors, an incremental carton and expense prevention opportunity (ICEPO) value for each of the one or more sets of unit identifiers based on a comparison of fulfillment of a set of unit identifiers from a reference fulfillment center using a single carton with fulfillment of a first unit identifier of the set of unit identifiers using a first carton and fulfillment of a second unit identifier of the set of unit identifiers using a second carton;
calculating, by the one or more processors, an adjusted fulfillment center stocking score using the ICEPO value and a difference between a first contribution value and a second contribution value for each of the one or more sets of unit identifiers;
selecting, by the one or more processors, a subset of the one or more sets of unit identifiers based on the adjusted fulfillment center stocking score;
computing, by the one or more processors, forecast data using a machine learning model trained using the order data;
generating, by the one or more processors, an electronic data signal based on the selected subset of the one or more sets of unit identifiers and the computed forecast data; and
performing one or more of providing a graphical display using the electronic data signal and transmitting, by the one or more processors, the electronic data signal representing the subset of the one or more sets of unit identifiers to an inventory replenishment computing system via an application programming interface, the inventory replenishment computing system managing stocking of items uniquely identified by unit identifiers.
13. The system of claim 12 , wherein the operations further include determining, by the one or more processors, the one or more affinities between the items uniquely identified by the unit identifiers based on the order data.
14. The system of claim 13 , wherein determining the one or more affinities between the items uniquely identified by the unit identifiers based on the order data includes
determining, by the one or more processors, a threshold frequency for identifying sets of unit identifiers,
determining, by the one or more processors, groups of unit identifiers using the order data, each group of unit identifiers appearing within at least one individual order, and
determining, by the one or more processors, a quantity of individual orders in which each group of unit identifiers appears.
15. The system of claim 14 , wherein generating the one or more sets of unit identifiers based on the one or more affinities between the items uniquely identified by the unit identifiers includes
selecting, by the one or more processors, the one or more sets of unit identifiers from among the groups of unit identifiers based on the quantity of individual orders in which each group of unit identifiers appears and the threshold frequency for identifying sets of unit identifiers.
16. The system of claim 12 , wherein transmitting the electronic data signal to the inventory replenishment computing system includes communicating the electronic data signals to the inventory replenishment computing system via a computer interface, the inventory replenishment computing system configured to stock items in the reference fulfillment center based on the electronic data signal.
17. The system of claim 16 , wherein the inventory replenishment computing system is further configured to stock items in the reference fulfillment center based on a forecasted demand of the items for a geographic region served by the reference fulfillment center.
18. The system of claim 12 , wherein calculating the ICEPO value for the one or more sets of unit identifiers includes determining a delivery expense from the reference fulfillment center to a defined location for the items uniquely identified by the one or more sets of unit identifiers.
19. The system of claim 18 , wherein calculating the ICEPO value for the one or more sets of unit identifiers includes determining a delivery expense from an alternate fulfillment center to the defined location for the items uniquely identified by the one or more sets of unit identifiers.
20. The system of claim 12 , wherein selecting the subset of the one or more sets of unit identifiers based on the adjusted fulfillment center stocking score includes
sorting the one or more sets of unit identifiers into a hierarchy based on the adjusted fulfillment center stocking score, and
selecting the subset of the one or more sets of unit identifiers based on a threshold quantity of sets of unit identifiers and an excess stocking capacity of a fulfillment center.