US-12626224-B2 - System and a method for managing inventory

US12626224B2US 12626224 B2US12626224 B2US 12626224B2US-12626224-B2

Abstract

There are provided methods and systems for managing inventory of inventory items in a storage area, and for automatically carrying out an action in response to a change in at least one inventory item in the storage area.

Inventors

Meir Zohar
Moshe ROSENBLUM
Ofer Solomon
David KOZLOVSKY
NOAM KRONMAN

Assignees

FRESHUB LTD

Dates

Publication Date: 20260512
Application Date: 20240410

Claims (14)

1 . A method for managing inventory in a storage area associated with a user, the storage area housing at least one inventory item and having a plurality of sensors associated therewith, the method being performed by a processor, said method comprising: receiving an image input signal from an image sensor of said plurality of sensors; extracting features from said image input signal, and, based on said features, identifying a plurality of candidate inventory items in said storage area; receiving a second input signal from a second sensor of said plurality of sensors; using said second input signal, uniquely identifying one of said plurality of candidate inventory items as being a specific one of said plurality of inventory items in said image input signal, wherein each of said first image input signal and said second input signal is insufficient, on its own, for facilitating unique identification of said specific one of said plurality of inventory items; based on said image input signal and said second input signal, as well as on user-specific information learned over time using machine-learning techniques, identifying a change in said specific one of said plurality of inventory items; in response to said identifying said change, automatically purchasing said specific one of said plurality of inventory items; wherein the user is associated with at least one segment of users, the at least one segment of users including a plurality of users sharing at least one common characteristic with the user, at least one of said plurality of users being associated with a different inventory than the inventory of said user, wherein said uniquely identifying said specific one of said plurality of inventory items is further based on segment-specific information for said at least one segment of users, wherein said segment-specific information is learned over time using machine-learning techniques by a machine learning module, based on information received from said user and from other users in said segment of users, and wherein said segment-specific information is common to at least a majority of users in said at least one segment of users.
2 . The method of claim 1 , wherein said user-specific information learned over time includes at least one of: information relating to a consumption rate of said specific one of said plurality of inventory items; information relating to a specific location in the storage area of said specific one of said plurality of inventory items; information relating to a purchasing pattern of said specific one of said plurality of inventory items; and information relating to a use pattern of said specific one of said plurality of inventory items.
3 . The method of claim 1 , wherein said uniquely identifying further includes using data obtained from a data repository including information about inventory items to uniquely identify said specific one of said plurality of inventory items.
4 . The method of claim 1 , wherein said receiving said image input signal and said receiving said second input signal is responsive to at least one triggering event.
5 . The method of claim 1 , wherein said receiving said second input signal comprises receiving a non-image signal as said second input signal.
6 . The method of claim 1 , wherein said extracting features comprises extracting color features from said image input signal.
7 . The method of claim 1 , wherein said extracting features comprises extracting dimension features from said image input signal.
8 . A method for managing inventory in a storage area associated with a user, the storage area housing a plurality of inventory items and having a plurality of sensors associated therewith, the method being performed by a processor, the method comprising: receiving an image input signal from an image sensor of said plurality of sensors and a second input signal from a second sensor of said plurality of sensors; extracting features from said image input signal, and, based on said features, identifying a plurality of candidate inventory items in said storage area; using said second input signal, uniquely identifying one of said plurality of candidate inventory items as being a specific one of said plurality of inventory items in said image input signal; based on at least one of said image input signal and said second input signal, identifying a change in said specific one of said plurality of inventory items; and in response to said identifying said change automatically purchasing said specific one of said plurality of inventory items, said automatically purchasing occurring at a time of said identifying or being scheduled to occur at a future time at which said specific one of said plurality of inventory items is predicted to expire or to be consumed, wherein the user is associated with at least one segment of users, the at least one segment of users including a plurality of users sharing at least one common characteristic with the user, at least one of said plurality of users being associated with a different inventory than the inventory of said user, wherein said uniquely identifying said specific one of said plurality of inventory items is further based on segment-specific information for said at least one segment of users, wherein said segment-specific information is learned over time using machine-learning techniques by a machine learning module, based on information received from said user and from other users in said segment of users, and wherein said segment-specific information is common to at least a majority of users in said at least one segment of users.
9 . The method of claim 8 , wherein said identifying said change further comprises processing said image input signal and/or said second input signal together with user-specific information learned over time using machine-learning techniques to identify said change.
10 . The method of claim 8 , wherein said receiving said image input signal and said second input signal is responsive to at least one triggering event.
11 . The method of claim 8 , wherein said second input signal is selected from the group consisting of: a second image signal; a pressure signal; a time based signal; a calendar based signal; an optical signal; a radio frequency signal; a photospectrometry signal; a raman spectrometry signal; a material oscillation measurement signal; a magnetic resonance measurement signal; a kinetic wave decay signal; an ultrasonic signal; and a chemical signal.
12 . The method of claim 8 , wherein said extracting features comprises extracting color features from said image input signal.
13 . The method of claim 8 , wherein said extracting features comprises extracting dimension features from said image input signal.
14 . The method of claim 8 , wherein said receiving comprises receiving a non-image input signal as said second input signal.

Description

RELATED APPLICATIONS The present application is a continuation of U.S. patent application Ser. No. 17/320,307, filed May 14, 2021, which in turn is a continuation of U.S. patent application Ser. No. 15/814,966, filed Nov. 16, 2017, both being entitled “A system and a Method for Managing Inventory”. U.S. patent application Ser. No. 15/814,966 gains priority from U.S. Provisional Patent Application 62/423,338, filed 17 Nov. 2016 and entitled “A System and a Method for Tracking Inventory Using Sensors and Computer Vision Techniques”. All of U.S. application Ser. Nos. 17/320,307, 15/814,966 and Provisional Patent Application No. 62/423,338 are incorporated herein by reference as if fully set forth herein. FIELD AND BACKGROUND OF THE INVENTION The invention, in some embodiments, relates to the field of managing inventory, and more particularly to methods and systems for automatically tracking inventory and consumption rates of inventory items and carrying out one or more actions based on identified changes in inventory. In many homes and businesses, the quantity of different products, or inventory items, needs to be tracked in order to determine which items need to be replenished and should be purchased, and for which items there is a sufficient quantity, so as to ensure proper stocking of the home or business and to prevent waste. Such determination is often time consuming, and is prone to errors. The ability to gauge inventory levels automatically, without user interaction, is complex, as it requires an identification of a package of a product within a storage area and monitoring of a quantity of the product within the package. This requires an understanding of the dimensions and contents of packages, which have posed a significant challenge to developers to date. Additionally, one of the actions users perform when checking inventory, is determining the usability of the remaining product—for example, when users check whether they need to buy milk, they typically check not only how much milk they have, but also when that milk will expire, or how long that milk will be usable. As such, an automatic system must also determine whether or not the identified product is usable, or for how long the product will remain useable, in order to determine the effective inventory of the product. There is thus a need in the art for an inventory gauging system and method which automatically identifies an inventory item, gauges its effective inventory by determining the available quantity and the usability of that quantity of the inventory item, and carries out an action, such as purchasing the inventory item, reminding the user to discard an inventory item, or providing an advertisement for a corresponding inventory item, in response to identification of the effective inventory. SUMMARY OF THE INVENTION Some embodiments of the invention relate to methods and systems for managing inventory, such as identifying inventory items in a storage area, gauging an effective quantity or inventory of the inventory items or a change in the inventory items, and carrying out an action in response to such gauging. In accordance with a first aspect of the present invention, there is provided a method for managing inventory in a storage area of a user, the storage area housing at least one inventory item and having a plurality of sensors associated therewith, the method including: receiving at least a first input signal from a first sensor of the plurality of sensors and a second input signal from a second sensor of the plurality of sensors, the first sensor and the second sensor being sensors of different types and the first input signal and the second input signal being signals of different types;processing first and second signals, based on the first and second input signals, respectively, together with user-specific information learned over time using machine-learning techniques to identify a change in at least one inventory item in the storage area;based on the identified change, carrying out at least one action, the at least one action including at least one of: adding the at least one inventory item to an inventory replenishment list;removing the at least one inventory item from an inventory replenishment list;updating a data repository to reflect the change to the at least one inventory item;purchasing the at least one inventory item;predicting a time at which the at least one inventory item will expire;predicting a time at which the at least one inventory item will be consumed;providing to the user an advertisement for an inventory item corresponding to or useable instead of the at least one inventory item;providing to the user an indication that the at least one inventory item has expired; andproviding to the user a recommendation to discard or to add the at least one inventory item. In some embodiments of the first aspect, at least the first signal is selected from the group consisting of: a photospectrometry input signal;a ra