# US 12,007,745 B2 · Apparatus for predicting equipment damage

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 12,007,745 B2
- Document ID: US12007745B2
- Status: Granted
- Kind: B2
- Application number: 17/480,165
- Filed: 2021-09-21
- Published: 2022-01-06
- Granted: 2024-06-11
- Earliest claimed: 2019-03-22
- Assignees: ABB Schweiz AG
- Applicants: ABB Schweiz AG
- Inventors: Ido Amihai; Subanatarajan Subbiah; Arzam Muzaffar Kotriwala; Moncef Chioua
- Web page: https://patentgrep.com/patent/US12007745B2

## CPC

- G05B 19/4065: Monitoring tool breakage, life or condition
- G05B 23/024: {Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks}
- G05B 23/0283: {Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL] (preventive maintenance, i.e. planning maintenance according to the available resources without monitoring the system G06Q10/06)}
- G06F 18/23213: with fixed number of clusters, e.g. K-means clustering
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/045: Combinations of networks
- G06N 3/09: Supervised learning
- G06N 3/0985: Hyperparameter optimisation; Meta-learning; Learning-to-learn
- G05B 2219/37252: Life of tool, service life, decay, wear estimation
- G05B: CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS (systems for controlling or regulating non-electric variables G05D; systems for regulating electric or magnetic variables G05F; control devices or systems insofar as characterised by mechanical features only G05G)
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

An apparatus includes an input unit, a processing unit, and an output unit. The input unit is configured to provide the processing unit with sensor data for an item of equipment. The processing unit is configured to implement at least one machine learning algorithm, which has been trained on the basis of a plurality of calibration sensor data for the item of equipment. Training of the at least one machine learning algorithm includes processing the plurality of calibration sensor data to determine at least two clusters representative of different equipment states. The processing unit is configured to implement the at least one machine learning algorithm to process the sensor data to assign the sensor data to a cluster of the at least two clusters to determine an equipment state for the item of equipment. The output unit is configured to output the equipment state for the item of equipment.

## Claims

### Claim 1

An apparatus for predicting equipment damage, the apparatus comprising:
an input;
a processor; and
an output;
wherein the input is configured to provide the processor with sensor data for an item of equipment,
wherein the processor is configured to implement at least one machine learning algorithm,
wherein the at least one machine learning algorithm has been trained based on a plurality of calibration sensor data for the item of equipment,
wherein training of the at least one machine learning algorithm comprises:
processing the plurality of calibration sensor data to determine a plurality of associated key condition indicator data,
transforming the key condition indicator data using a moving average calculation over time to obtain transformed key condition indicator data, and
determining at least two clusters representative of different equipment states by determining at least two distributions of the transformed key condition indicator data,
wherein the processor is configured to implement the at least one machine learning algorithm to process the sensor data to assign the sensor data to a cluster of the at least two clusters to determine an equipment state for the item of equipment, and
wherein the output is configured to output the equipment state for the item of equipment.

### Claim 2 (depends on 1)

The apparatus according to claim 1, wherein: the moving average calculation is performed using an equation V(t)=β*V(t−1)+(1−β)*θ(t), wherein V(t) is a value of the moving average for a current time t, V(t−1) is a value of the moving average for a previous time t−1, β is a decay parameter, and θ(t) is an input at time t.

### Claim 3 (depends on 1)

The apparatus according to claim 1, wherein the determination of the at least two distributions of the transformed key condition indicator data is based on at least two peaks in the transformed key condition indicator data.

### Claim 4 (depends on 1)

The apparatus according to claim 1, wherein:
the sensor data are velocity data and the calibration sensor data are velocity data, and
the associated key condition indicator data are root mean square velocity data.

### Claim 5 (depends on 4)

The apparatus according to claim 4, wherein the sensor data are rotational velocity data, and the calibration sensor data are rotational velocity data.

### Claim 6 (depends on 1)

The apparatus according to claim 1, wherein training of the at least one machine learning algorithm comprises utilization of user input data relating to an equipment state associated with at least one subset of the calibration data.

### Claim 7 (depends on 6)

The apparatus according claim 6, wherein training of the at least one machine learning algorithm comprises utilization of user input data relating to a different equipment state associated with at least one second subset of the calibration data.

### Claim 8 (depends on 1)

A system for predicting equipment damage, the system comprising:
at least one sensor; and
the apparatus for predicting equipment damage according to claim 1,
wherein the at least one sensor is configured to acquire the sensor data and provide the sensor data to the processor.

### Claim 9 (depends on 8)

The system according to claim 8, wherein the at least one sensor is configured to acquire the calibration sensor data.

### Claim 10

A method for predicting equipment damage, the method comprising:
training at least one machine learning algorithm based on a plurality of calibration sensor data for an item of equipment, wherein training of the at least one machine learning algorithm comprises:
processing the plurality of calibration sensor data to determine a plurality of associated key condition indicator data,
transforming the key condition indicator data using a moving average calculation over time to obtain transformed key condition indicator data, and
determining at least two clusters representative of different equipment states by determining at least two distributions of the transformed key condition indicator data;
implementing by a processor the at least one machine learning algorithm to process sensor data for the item of equipment to assign the sensor data to a cluster of the at least two clusters to determine an equipment state for the item of equipment; and
outputting the equipment state for the item of equipment.

### Claim 11

A non-transitory computer-readable medium that stores a computer program, which upon execution by a processor, causing performance of a method, the method comprising:
training at least one machine learning algorithm based on a plurality of calibration sensor data for an item of equipment, wherein the training of the at least one machine learning algorithm comprises:
processing the plurality of calibration sensor data to determine a plurality of associated key condition indicator data,
transforming the key condition indicator data using a moving average calculation over time to obtain transformed key condition indicator data, and
determining at least two clusters representative of different equipment states by determining at least two distributions of the transformed key condition indicator data;
implementing the at least one machine learning algorithm to process sensor data for the item of equipment to assign the sensor data to a cluster of the at least two clusters to determine an equipment state for the item of equipment; and
outputting the equipment state for the item of equipment.