Predictive inventory availability

Patent metadata

Field	Value
Publication number	US-11544810-B2
Application number	US-201815885492-A
Country	US
Kind code	B2
Filing date	Jan 31, 2018
Priority date	Jan 31, 2018
Publication date	Jan 3, 2023
Grant date	Jan 3, 2023

Abstract

A method for predicting inventory availability, involving receiving a delivery order including a plurality of items and a delivery location, and identifying a warehouse for picking the plurality of items. The method retrieves a machine-learned model that predicts a probability that an item is available at the warehouse. The machine-learned model is trained, using machine learning, based in part on a plurality of datasets. The plurality of datasets include data describing items included in previous delivery orders, whether each item in each previous delivery order was picked, a warehouse associated with each previous delivery order, and a plurality of characteristics associated with each of the items. The method predicts the probability that one of the plurality of items in the delivery order is available at the warehouse, and generates an instruction to a picker based on the probability. An instruction is transmitted to a mobile device of the picker.

First claim

What is claimed is:

1. A method for predicting inventory availability, the method comprising:
receiving a delivery order comprising a plurality of items and a delivery location;
identifying a warehouse for picking the plurality of items based on the plurality of items and the delivery location;
training a machine-learned model on a set of training data, the set of training data describing items included in previous delivery orders, a warehouse associated with each previous delivery order, and a plurality of characteristics associated with each of the items, wherein the training comprises:
inputting factors for an item-warehouse pair into the machine-learned model, the item-warehouse pair corresponding to a specific item, wherein the machine-learned model generates a confidence score indicative of an accuracy of a prediction of availability of the specific item;
in response to the confidence score indicative of the accuracy of the prediction of availability of the specific item being below a threshold, sending instructions to a first mobile device of a first picker, the instructions comprising having the first picker to attempt to find the specific item and to collect new information about the item;
receiving, from the first mobile device, a result of the attempt to find the specific item and new information about the item;
updating the training data to include the new information, wherein the new information is used as the factors of the item-warehouse pair and wherein the result of the attempt to find the specific item is used to update the training data to include an indication of the availability of the specific item in the warehouse and a corresponding time of the result; and
training the machine-learned model on the updated training data;

predicting, using the machine-learned model, a probability that one of the plurality of items in the delivery order is available at the warehouse;
generating an instruction to a second picker based on the probability; and
transmitting the instruction to a second mobile device of the second picker.

2. The method of claim 1 , further comprising:
updating the training data at a periodic interval; and
retraining the machine-learned model based on updated training data.

3. The method of claim 1 , wherein the training data further comprises data identifying, for each previously-picked delivery order, a time associated with the previously-picked delivery order, and wherein the time comprises at least one of: a time of day the delivery order is picked, a day of a week the delivery order is picked, a time interval since the item was picked in a previously-picked delivery order, and a time interval since the item was not found in a previous delivery order.

4. The method of claim 1 , wherein the plurality of characteristics associated with an item comprises at least one of: a department associated with the item, an aisle of the warehouse associated with the item, an item popularity score, a product type associated with the item, a time interval since the item was found, and a time interval since the item was not found.

5. The method of claim 1 , wherein generating the instruction to the second picker based on the probability comprises:
receiving an indication from the second picker that the second picker cannot find the item;
determining that the probability for an item of the plurality of items in the delivery order is above a threshold that indicates that the item is available at the warehouse;
instructing the second picker to continue looking for the item.

6. The method of claim 5 , further comprising:
providing a location within the warehouse that the item is most likely available.

7. The method of claim 1 , further comprising:
generating a warning to a user associated with the delivery order if the probability is below a threshold value.

8. The method of claim 1 , wherein the warehouse is one of a plurality of potential warehouses, and the warehouse is selected from the plurality of potential warehouses based on the probability that one of the plurality of items in the delivery order is available at the warehouse.

9. A non-transitory computer-readable storage medium storing instructions for predicting inventory availability, wherein the instructions, when executed by a processor, cause the processor to:
receive a delivery order comprising a plurality of items and a delivery location;
identify a warehouse for picking the plurality of items based on the plurality of items and the delivery location;
train a machine-learned model on a set of training data, the set of training data describing items included in previous delivery orders, a warehouse associated with each previous delivery order, and a plurality of characteristics associated with each of the items, wherein the training comprises:
input factors for an item-warehouse pair into the machine-learned model, the item-warehouse pair corresponding to a specific item, wherein the machine-learned model generates a confidence score indicative of an accuracy of a prediction of availability of the specific item;
in response to the confidence score indicative of the accuracy of the prediction of availability of the specific item being below a threshold, send instructions to a first mobile device of a first picker, the instructions comprising having the first picker to attempt to find the specific item and to collect new information about the item;
receive, from the first mobile device, a result of the attempt to find the specific item and new information about the item;
update the training data to include the new information, wherein the new information is used as the factors of the item-warehouse pair and wherein the result of the attempt to find the specific item is used to update the training data to include an indication of the availability of the specific item in the warehouse and a corresponding time of the result; and
train the machine-learned model on the updated training data;

predict, using the machine-learned model, a probability that one of the plurality of items in the delivery order is available at the warehouse;
generate an instruction to a second picker based on the probability; and
transmit the instruction to a second mobile device of the second picker.

10. The computer-readable storage medium of claim 9 , wherein the instructions, when executed, further cause the processor to:
update the training data at a periodic interval; and
retrain the machine-learned model based on updated training data.

11. The computer-readable storage medium of claim 9 , wherein the training data further comprises data identifying, for each previously-picked delivery order, a time associated with the previously-picked delivery order, and wherein the time comprises at least one of: a time of day the delivery order is picked, a day of a week the delivery order is picked, a time interval since the item was picked in a previously-picked delivery order, and a time interval since the item was not found in a previous delivery order.

12. The computer-readable storage medium of claim 9 , wherein the plurality of characteristics associated with an item comprises at least one of: a department associated with the item, an aisle of the warehouse associated with the item, an item popularity score, a product type associated with the item, a time interval since the item was found, and a time interval since the item was not found.

13. The computer-readable storage medium of claim 9 , wherein generate the instruction to the second picker based on the probability comprises:
receive an indication from the second picker that the second picker cannot find the item;
de

Assignees

• Maplebear Inc

Inventors

• Rao Sharath
• Prasad Shishir
• Stanley Jeremy

Classifications

• G06Q10/087 Primary

  Inventory or stock management, e.g. order filling, procurement or balancing against orders · CPC title

• G06Q10/06315

  Needs-based resource requirements planning or analysis · CPC title

• G06N5/01

  Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound · CPC title

• G06Q10/067

  Enterprise or organisation modelling · CPC title

• G06N20/20

  Ensemble learning · CPC title