Machine learning techniques for directed storage guidance in inventory bins of moveable inventory holders
A method may include determining inventory item information associated with an inventory item to be stowed in an inventory system that comprises a plurality of moveable inventory holders. The method may also include determining inventory holder information that characterizes one or more properties of a moveable inventory holder of the plurality of moveable inventory holders. The moveable inventory holder may include a plurality of inventory bins. The method may also include determining, using a machine learning model, a set of candidate inventory bins of the plurality of inventory bins based on the inventory item information and the inventory holder information. The method may also include providing for presentation a set of cues corresponding to the set of candidate inventory bins.
1. An inventory system, comprising:
a plurality of moveable inventory holders;
a projection system;
a scanner system; and
a computer system operably coupled with the projection system and the scanner system, the computer system configured to:
receive, from the scanner system, scan information that identifies an inventory item to be stowed in a moveable inventory holder of the plurality of moveable inventory holders;
responsive to receiving the scan information, determine inventory item information associated with the inventory item;
determine inventory holder information that characterizes one or more inventory holder properties of a moveable inventory holder of the plurality of moveable inventory holders, the moveable inventory holder comprising a plurality of inventory bins;
determine, using a machine learning model, a set of candidate inventory bins of the plurality of inventory bins using the inventory item information and the inventory holder information as input into the machine learning model;
present, using the projection system, a set of visual cues corresponding to the set of candidate inventory bins;
register placement of the inventory item into a candidate bin of the set of candidate inventory bins; and
update the inventory holder information based at least in part on the placement.
2. The inventory system of claim 1 , wherein the machine learning model models historical stow behavior of a set of historical users who stowed historical inventory items into historical inventory bins of historical moveable inventory holders.
3. The inventory system of claim 1 , wherein the one or more inventory holder properties represent a current state of the moveable inventory holder.
4. The inventory system of claim 3 , wherein the current state is dependent on a distribution of current inventory items currently stowed among the plurality of inventory bins.
5. The inventory system of claim 1 , wherein determining, using the machine learning model, the set of candidate inventory bins comprises:
computing a score for each inventory bin of the plurality of inventory bins;
ranking each inventory bin of the plurality of inventory bins using a respective score to define a ranked set of inventory bins; and
selecting the set of candidate inventory bins by applying a threshold to the ranked set of inventory bins.
6. The inventory system of claim 5 , wherein the threshold comprises a score-based threshold or a predefined numerical quantity threshold.
7. A computer-implemented method, comprising:
determining inventory item information associated with an inventory item to be stowed in an inventory system that comprises a plurality of moveable inventory holders;
determining inventory holder information that characterizes one or more properties of a moveable inventory holder of the plurality of moveable inventory holders, the moveable inventory holder comprising a plurality of inventory bins;
determining, using a machine learning model, a set of candidate inventory bins of the plurality of inventory bins based on the inventory item information and the inventory holder information;
providing for presentation a set of cues corresponding to the set of candidate inventory bins;
registering placement of the inventory item into a candidate bin of the set of candidate inventory bins; and
updating the inventory holder information based at least in part on the placement.
8. The computer-implemented method of claim 7 , wherein the set of cues are visual cues.
9. The computer-implemented method of claim 8 , providing for presentation the set of cues comprises providing the set of cues for projection onto one side of the moveable inventory holder that includes the plurality of inventory bins.
10. The computer-implemented method of claim 9 , further comprising:
determining a set of excluded candidate inventory bins that are distinct from the set of candidate inventory bins; and
providing a set of negative hints for projection onto the one side of the moveable inventory holder, the set of negative hints identifying the set of excluded candidate inventory bins.
11. The computer-implemented method of claim 7 , wherein determining inventory holder information comprises receiving a set of precomputed bin rankings that rank the plurality of inventory bins of the moveable inventory holder.
12. The computer-implemented method of claim 7 , wherein the inventory item information comprises item attributes of the inventory item.
13. The computer-implemented method of claim 7 , further comprising receiving scan information that identifies the inventory item, and wherein determining the inventory item information is responsive to receiving the scan information.
14. The computer-implemented method of claim 7 , further comprising:
determining new inventory item information associated with a new inventory item;
determining, using the machine learning model, a new set of candidate inventory bins of the plurality of inventory bins based on the new inventory item information and the updated inventory holder information; and
providing for presentation a new set of cues corresponding to the new set of candidate inventory bins.
15. One or more non-transitory computer-readable media comprising computer-executable instructions that, when executed by one or more processors of a computer system, cause the computer system to perform operations comprising:
determining inventory item information associated with an inventory item to be stowed in an inventory system that comprises a plurality of moveable inventory holders;
determining inventory holder information that characterizes one or more properties of a moveable inventory holder of the plurality of moveable inventory holders, the moveable inventory holder comprising a plurality of inventory bins;
generating, using a machine learning model, candidate inventory bin information that represents at least one candidate inventory bin of the plurality of inventory bins based on the inventory item information and the inventory holder information;
providing a visual cue for presentation on the moveable inventory holder, the visual cue identifying the at least one candidate inventory bin;
registering placement of the inventory item into the at least one candidate bin; and
updating the inventory holder information based at least in part on the placement.
16. The one or more non-transitory computer-readable media of claim 15 , wherein the machine learning model outputs relevancy scores between the inventory item and each of the plurality of inventory bins.
17. The one or more non-transitory computer-readable media of claim 16 , wherein generating the candidate inventory bin information comprises generating the candidate inventory bin information based at least in part on the relevancy scores for a subset of the plurality of inventory bins.
18. The one or more non-transitory computer-readable media of claim 15 , wherein determining the inventory holder information comprises receiving a set of precomputed bin rankings that rank the plurality of inventory bins of the moveable inventory holder.