Managing a removal channel based on inventory capacity

• US 9,818,082 B1
• Filed: 01/28/2016
• Issued: 11/14/2017
• Est. Priority Date: 01/28/2016
• Status: Expired due to Fees

• Current Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Original Assignee: Amazon Technologies, Inc. (Amazon.com, Inc.)
• Inventors: Sadighian, Ali, Humair, Salal, Tong, Jun, Hu, Xiaochang, Bishr, Mahmoud Ahmed, Liu, Chao, Wu, Di, Brayak, Travis Francis, Bajari, Patrick Ludvig, Bhatia, Deepak
• Primary Examiner(s): Ade, Garcia
• Application Number:
  US15/009,607
• Time in Patent Office: 656 Days
• Field of Search: 705 28
• CPC Class Codes: G06Q 10/087 Inventory or stock management, e.g. order filling, procurement or balancing against orders

Abstract

Techniques for managing a removal channel may be described. In an example, a first estimate may be generated. The first estimate may be associated with removing a volume of items from an inventory space within a planning horizon through a removal channel. A decision to remove the volume of items may be accessed. The decision may be based on the first estimate and a second estimate associated with adding a capacity for inventorying an equivalent volume of items. Based on the decision, a constraint may be generated and imposed on an inventory level of an item during a time period of the planning horizon. A quantity of the item to remove within the time period through the removal channel may be estimated based on the constraint. Usage of the removal channel during the time period may be set based on the quantity of the item to remove.

Claims

1. A computer-implemented method comprising:

generating, by a computing system based at least in part on a first model, a first estimate associated with removing a first volume of items from an inventory space during a planning horizon through a removal channel;

accessing, by the computing system, a decision to remove the volume of items, the decision based at least in part on the first estimate and a second estimate, the second estimate associated with adding a space capacity for inventorying a second volume of items that is equivalent to the first volume, the second estimate generated by a second computing system based at least in part on a second model that is different than the first model, the second estimate received by the computing system from the second computing system based at least in part on a transmission of the first estimate from the computing system to the second computing system;

generating, by the computing system based at least in part on the decision, a constraint on an inventory level of an item during a next time period of a plurality of time periods that form the planning horizon, the constraint based at least in part on the first volume of the items to be removed during the planning horizon; and

estimating, by the computing system, a quantity of the item to remove within the next time period through the removal channel, the quantity estimated based at least in part on a search that uses an objective function, the objective function using the constraint and a number of variables to increase or decrease an optimization parameter, the objective function using the quantity of the item as one of the variables,

wherein usage of the removal channel during the next time period is controlled based at least in part on the quantity of the item, wherein the usage comprises physical removal of units of the item from the inventory space.

2. The computer-implemented method of claim 1, wherein generating the first estimate comprises generating a smallest possible loss of removing the first volume of items as part of removing inventoried items through a plurality of removal channels of an electronic marketplace associated with the inventory space, and wherein the plurality of removal channels comprise a markdown channel of the electronic marketplace.

3. The computer-implemented method of claim 1, wherein the constraint comprises a penalty that is set based at least in part on the first estimate, wherein estimating the quantity of the item to remove comprises increasing or decreasing the optimization parameter based at least in part on the penalty in the objective function and setting the quantity of the item to remove as one of the variables.

4. The computer-implemented method of claim 3, wherein generating the first estimate comprises varying the penalty within a range and using the objective function to find the first estimate, and wherein the penalty is set as a value within the range corresponding to finding the first estimate.

5. The computer-implemented method of claim 3, further comprising:

upon an elapse of the next time period, accessing information about an actual quantity of the item removed within the next time period;

updating the penalty based at least in part on the actual quantity; and

estimating a next quantity of the item to remove within a subsequent time period based at least in part on the penalty being updated.

6. The computer-implemented method of claim 5, wherein the penalty varies within a range, and wherein updating the penalty comprises increasing or decreasing the optimization parameter based at least in part on a bisection search within the range.

7. The computer-implemented method of claim 1, wherein estimating the quantity based at least in part on the search comprises recursively estimating quantities of the item to remove in future time periods of the planning horizon, based at least in part on the constraint.

8. The computer-implemented method of claim 1, further comprising estimating a removal value associated with removing the quantity of the item within the next time period.

9. The computer-implemented method of claim 8, wherein the removal channel comprises a markdown channel, wherein controlling the usage of the markdown channel is based at least in part on the quantity to remove through the markdown channel within the next time period and based at least in part on the removal value.

16. The computer-implemented method of claim 1, wherein estimating the quantity of the item to remove comprises based at least in part on the search comprises recursively searching for the quantity by at least starting with a last time period occurring at the end of the planning horizon, moving through one or more time periods of the planning horizon, and ending at the next time period, and using the objective function in each of the time periods based at least in part on the updated constraint, the variables, and the optimization parameter, and wherein an updated constraint in a particular time period is set based at least in part on the recursive searching in a subsequent time period.

10. A computing system, comprising:

one or more processors;

one or more computer-readable media comprising instructions that, upon execution with the one or more processors, cause the computing system to;

generate, based at least in part on a first model, a first estimate for removing a volume of items through a plurality of removal channels to free a space capacity of an inventory space over a planning horizon;

receive a decision to remove the volume of items, the decision based at least in part on the first estimate and a second estimate, the second estimate associated with adding an inventory capacity that is equivalent to the space capacity to be freed, the second estimate generated by a second computing system based at least in part on a second model that is different than the first model, the second estimate received by the computing system from the second computing system based at least in on a transmission of the first estimate from the computing system to the second computing system;

generate, based at least in part on the decision, a penalty associated with removing less than the volume of items over the planning horizon; and

estimate a quantity of an inventoried item to be removed during a next time period of a plurality of time periods that form the planning horizon through a removal channel of the plurality of removal channels, the quantity estimated based at least in part on a search that uses an objective function, the objective function using the penalty as a variable and a number of other variables to increase or decrease an optimization parameter, the objective function using the quantity of the item as one of the other variables,

wherein at least one parameter associated with using the removal channel during the next time period is controlled based at least in part on the quantity of the inventoried item, wherein using the removal channel comprises physical removal of units of the item from the inventory space.

11. The computing system of claim 10, wherein the instructions, upon execution with the one or more processors, further cause the computing system to:

provide the quantity of the inventoried item to be removed to a third computing system that is configured to generate a plan for the physical removal of units of the item during the next time period.

12. The computing system of claim 11, wherein the removal channel comprises a markdown channel, wherein the at least one parameter is part of the plan and comprises a markdown price of the item, and wherein the instructions, upon execution with the one or more processors, further cause the computing system to:

generate a removal value associated with removing the inventoried item;

provide the removal value to the third computing system, wherein the plan is based at least in part on the quantity of the inventoried item to be removed and based at least in part on the removal value.

13. The computing system of claim 10, wherein the instructions, upon execution with the one or more processors, further cause the computing system to:

determine an actual quantity of the inventoried item removed by an end of the next time period;

update the penalty based at least in part on the actual quantity; and

estimate a next quantity of the inventoried item to be removed in a subsequent time period of the planning horizon, wherein the next quantity is recursively estimated based at least in part on the penalty and based at least in part on remaining subsequent time periods of the planning horizon.

14. The computing system of claim 10, wherein the decision is received from the second computing system that is configured to generate a plan to add the inventory capacity, and wherein the first estimate is updated and provided to the second computing system on a periodic basis.

15. The computing system of claim 10, wherein at least one of generating the first estimate or estimating the quantity of the inventoried item are performed based at least in part on a request from another computing system, wherein the request is received based at least in part on an application programming interface.