IP Library › Granted Patent US 12693272

Granted Patent B2

US 12693272 · App. 18/106,669 · Granted Jul 28, 2026

Door fault identification

Inventors: Daniel Langenberg (Zionsville, IN); Nathanael L. Thomas (Carmel, IN)

Assignee: Schlage Lock Company LLC

G01N29/4454 G01C19/5776 G01N29/14 G01N29/4481 G01P13/00 G01P15/0802 G06N3/08 G01N2291/0258

View Patent ↗

Loading inventors, assignments & file history…

Monitor This Case

Get email alerts when status or documents change.

Order Certified Copies

Most orders are placed with the USPTO same day — all within 24 business hours.

Order via The Patent Place →

Pre-filled with this patent's details

Quick Facts

Patent No.: US 12693272
App. No.: 18/106,669
Granted: Jul 28, 2026
Kind: B2

External Links

USPTO Patent Center↗ Full Patent PDF↗

Abstract

A method according to one embodiment includes receiving sensor data from a plurality of sensors of a door device associated with a door, analyzing the sensor data to determine behavior data indicative of a behavior of the door device, and comparing the behavior data to a plurality of representative data associated with a plurality of door faults to determine a corresponding likelihood that the sensor data corresponds with each of the door faults.

Claims (95)

1 . An auto-operator, comprising:

a plurality of sensors configured to generate sensor data associated with one or more characteristics of a door;

at least one processor; and

at least one memory comprising a plurality of instructions stored thereon that, in response to execution by the at least one processor, causes the auto-operator to:

receive the sensor data from the plurality of sensors;

analyze the sensor data to determine behavior data indicative of a behavior of the auto-operator; and

compare the behavior data to a plurality of representative data associated with a plurality of door faults to determine a corresponding likelihood that the sensor data corresponds with each of the door faults;

wherein the at least one memory further comprising an artificial neural network and corresponding weights, and wherein the plurality of instructions further causes the auto-operator to:

receive input indicative of maintenance performed on the door;

receive new sensor data from the plurality of sensors in response to receipt of the input indicative of the maintenance performed;

analyze the new sensor data to determine new behavior data indicative of the behavior of the auto-operator;

compare the new behavior data to the plurality of representative data associated with the plurality of door faults to determine a new corresponding likelihood that the new sensor data corresponds with each of the door faults based on the artificial neural network and the corresponding weights; and

update the weights of the artificial neural network based on the maintenance performed and the comparison of the new behavior data to the plurality of representative data, wherein each of the weights is associated with a corresponding door fault.

2 . The auto-operator of claim 1 , wherein the at least one memory comprises a fault prediction database that includes the data associated with door faults and the corresponding weights of the artificial neural network;

wherein the plurality of instructions further causes the auto-operator to determine whether the fault prediction database associates the maintenance performed on the door as a solution for a likely fault associated with the new sensor data; and

wherein to update the corresponding weights of the artificial neural network in the fault prediction database comprises to update the corresponding weights in response to a determination that the fault prediction database already associates the maintenance performed on the door as a solution for the likely fault associated with the new sensor data.

3 . The auto-operator of claim 2 , wherein the plurality of instructions further causes the auto-operator to:

compare the maintenance performed on the door to an unverified fault prediction database of unverified solutions; and

update the corresponding weights of the artificial neural network in the unverified fault prediction database in response to a match for the maintenance performed in the unverified fault prediction database;

wherein possible solutions to faults in the unverified fault prediction database are presented as possible solutions only if an associated weight of the possible solution exceeds a predetermined threshold.

4 . The auto-operator of claim 1 , further comprising a primary housing that includes the at least one processor and the at least one memory; and

wherein at least one of the plurality of sensors is located external to the primary housing.

5 . The auto-operator of claim 1 , wherein a first sensor of the plurality of sensors is configured to detect a force on the door and wake other sensors of the plurality of sensors in response to detection of the force on the door.

6 . The auto-operator of claim 1 , wherein to compare the behavior data to the plurality of representative data associated with the plurality of door faults comprises to segment the behavior data into a plurality of operating regions, wherein each operating region of the plurality of operating regions is associated with a corresponding predefined door zone, and wherein each corresponding predefined door zone is associated with a range of door angles relative to a closed position of the door.

7 . The auto-operator of claim 6 , wherein to segment the behavior data into the plurality of operating regions comprises to segment the behavior data into at least a first operating region associated with an opening zone, a second operating region associated with a hold-open zone, and a third operating region associated with a closing zone.

8 . The auto-operator of claim 6 , wherein to segment the behavior data into the plurality of operating regions comprises to segment the behavior data into a first operating region associated with an initial opening zone, a second operating region associated with a latch zone during opening, a third operating region associated with a main zone during opening, a fourth operating region associated with a hold-open zone, a fifth operating region associated with a main zone during closing, a sixth operating region associated with a latch zone during closing, a seventh operating region associated with a frame zone during closing, an eighth operating region associated with a closed position, and a ninth operating region associated with a resting position.

9 . The auto-operator of claim 6 , wherein to compare the behavior data to the plurality of representative data associated with the plurality of door faults comprises to:

select a first feature;

select a first operating region from the plurality of operating regions; and

perform feature detection to the behavior data within the first operating region to generate a first feature detection score based on the first feature.

10 . The auto-operator of claim 9 , wherein to compare the behavior data to the plurality of representative data associated with the plurality of door faults further comprises to:

select a second feature;

perform feature detection to the behavior data within the first operation region to generate a second feature detection score based on the second feature; and

generate an operating region score based on the first feature detection score and the second feature detection score.

11 . The auto-operator of claim 10 , wherein to compare the behavior data to the plurality of representative data associated with the plurality of door faults further comprises to generate one or more system scores based on the operating region score and at least one other operating region score associated with a second operating region selected from the plurality of operating regions; and

wherein each of the one or more system scores comprises an output value of the artificial neural network.

12 . The auto-operator of claim 1 , wherein the plurality of instructions further causes the auto-operator to:

detect, by a first sensor of the plurality of sensors, movement of the door; and

transition other sensors of the plurality of sensors from a low power sleep state to a higher power state in response to detecting movement of the door.

13 . The auto-operator of claim 1 , wherein the plurality of sensors further comprises one or more environmental sensors adapted to generate sensor data indicative of a physical environment of the auto-operator;

wherein to analyze the sensor data comprises to (i) at least one of filter or synthesize the sensor data to determine one or more inertial measurements indicative of the behavior of the auto-operator, and (ii) determine an environmental context of the auto-operator based on the sensor data generated by the one or more environmental sensors; and

wherein the behavior data is determined based on the one or more inertial measurements and the environmental context of the auto-operator.

14 . A door fault identification system, comprising:

at least one processor; and

at least one memory comprising data associated with door faults, an artificial neural network and corresponding weights of the artificial neural network, and a plurality of instructions stored thereon that, in response to execution by the at least one processor, causes the door fault identification system to:

receive sensor data from a plurality of sensors of an electronic lock associated with one or more characteristics of a door;

analyze the sensor data with the artificial neural network to determine behavior data indicative of a behavior of the electronic lock; and

compare, with the artificial neural network, the behavior data to a plurality of representative data associated with a plurality of door faults to determine a corresponding likelihood that the sensor data corresponds with each of the door faults based on the artificial neural network;

wherein the plurality of instructions further causes the door fault identification system to:

receive input indicative of maintenance performed on the door;

receive new sensor data from the plurality of sensors in response to receipt of the input indicative of the maintenance performed;

analyze the new sensor data to determine new behavior data indicative of the behavior of the electronic lock;

15 . The door fault identification system of claim 14 , wherein the at least one memory comprises a fault prediction database that includes the data associated with door faults and the weights of the artificial neural network;

wherein the plurality of instructions further causes the door fault identification system to determine whether the fault prediction database associates the maintenance performed on the door as a solution for a likely fault associated with the new sensor data; and

wherein to update the weights of the artificial neural network in the fault prediction database comprises to update the weights in response to a determination that the fault prediction database already associates the maintenance performed on the door as a solution for the likely fault associated with the new sensor data.

16 . The door fault identification system of claim 15 , wherein the plurality of instructions further causes the door fault identification system to:

compare the maintenance performed on the door to an unverified fault prediction database of unverified solutions; and

update weights of the artificial neural network in the unverified fault prediction database in response to a match for the maintenance performed in the unverified fault prediction database;

17 . The door fault identification system of claim 14 , wherein to compare the behavior data to the plurality of representative data associated with the plurality of door faults comprises to segment the behavior data into a plurality of operating regions, wherein each operating region of the plurality of operating regions is associated with a corresponding predefined door zone, and wherein each corresponding predefined door zone is associated with a range of door angles relative to a closed position of the door.

18 . The door fault identification system of claim 17 , wherein to segment the behavior data into the plurality of operating regions comprises to segment the behavior data into a first operating region associated with an initial opening zone, a second operating region associated with a latch zone during opening, a third operating region associated with a main zone during opening, a fourth operating region associated with a hold-open zone, a fifth operating region associated with a main zone during closing, a sixth operating region associated with a latch zone during closing, a seventh operating region associated with a frame zone during closing, an eighth operating region associated with a closed position, and a ninth operating region associated with a resting position.

19 . The door fault identification system of claim 17 , wherein to compare the behavior data to the plurality of representative data associated with the plurality of door faults comprises to:

select a first feature;

select a first operating region from the plurality of operating regions;

perform feature detection to the behavior data within the first operating region to generate a first feature detection score based on the first feature;

select a second feature;

perform feature detection to the behavior data within the first operation region to generate a second feature detection score based on the second feature;

generate an operating region score based on the first feature detection score and the second feature detection score;

generate one or more system scores based on the operating region score and at least one other operating region score associated with a second operating region selected from the plurality of operating regions; and

wherein each of the one or more system scores comprises an output value of the artificial neural network.

20 . The door fault identification system of claim 14 , wherein the plurality of instructions further causes the door fault identification system to:

detect, by a first sensor of a plurality of sensors of the electronic lock, movement of the door; and

transition other sensors of the plurality of sensors of the electronic lock from a low power sleep state to a higher power state in response to detecting movement of the door.

21 . An auto-operator, comprising:

a plurality of sensors configured to generate sensor data associated with one or more characteristics of a door;

at least one processor; and

receive the sensor data from the plurality of sensors;

analyze the sensor data to determine behavior data indicative of a behavior of the auto-operator; and

wherein the plurality of instructions further causes the auto-operator to:

receive input indicative of maintenance performed on the door;

receive new sensor data from the plurality of sensors in response to receipt of the input indicative of the maintenance performed;

analyze the new sensor data to determine new behavior data indicative of the behavior of the auto-operator;

22 . The auto-operator of claim 21 , wherein the at least one memory comprises a fault prediction database that includes the data associated with door faults and the weights of the artificial neural network;

23 . The auto-operator of claim 22 , wherein the plurality of instructions further causes the auto-operator to:

compare the maintenance performed on the door to an unverified fault prediction database of unverified solutions; and

update weights of the artificial neural network in the unverified fault prediction database in response to a match for the maintenance performed in the unverified fault prediction database;