# US 11,442,444 B2 · System and method for forecasting industrial machine failures > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,442,444 B2 - Document ID: US11442444B2 - Status: Granted - Kind: B2 - Application number: 17/167,243 - Filed: 2021-02-04 - Published: 2021-05-27 - Granted: 2022-09-13 - Earliest claimed: 2018-08-12 - Assignees: Aktiebolaget Skf - Applicants: Aktiebolaget SKF - Inventors: David Lavid Ben Lulu; Nir Dromi - Web page: https://patentgrep.com/patent/US11442444B2 ## CPC - 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)} - G06N 3/08: Learning methods - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/09: Supervised learning - G06N 3/0985: Hyperparameter optimisation; Meta-learning; Learning-to-learn - G06N 20/00: Machine learning - G05B 19/4184: {characterised by fault tolerance, reliability of production system} - G05B 23/0254: {based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks} - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 3/045: Combinations of networks - G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound - G06N 7/01: Probabilistic graphical models, e.g. probabilistic networks - Y02P 90/02: Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS] - 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) - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - Y02P: CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS ## Abstract A system and method for forecasting failures in industrial machines, including: receiving raw sensory inputs collected from at least one machine; generating a plurality of data features based on the raw sensory inputs; selecting from the plurality of data features a plurality of indicative data features, wherein the selection is based on a distribution of the plurality of indicative data features that determines an association between the plurality of indicative data features and a machine failure; selecting, based on the plurality of indicative data features, a machine learning model; applying the selected machine learning model to the plurality of indicative data features; and determining a probability for a forthcoming machine failure. ## Claims ### Claim 1 A method for forecasting failures in industrial machines, comprising: receiving raw sensory inputs collected from at least one machine; generating a plurality of data features based on the raw sensory inputs; selecting from the plurality of data features a plurality of indicative data features, wherein the selection is based on a distribution of the plurality of indicative data features that determines an association between the plurality of indicative data features and a machine failure, wherein the distribution of the plurality of indicative data features is a set of abnormal parameters of each of the indicative data features; selecting, based on the plurality of indicative data features, a machine learning model; applying the selected machine learning model to the plurality of indicative data features; and determining a probability for a forthcoming machine failure. ### Claim 2 (depends on 1) The method of claim 1, wherein each type of the raw sensory inputs is related to at least a process that is associated with the machine. ### Claim 3 (depends on 1) The method of claim 1, further comprising: preprocessing the received raw sensory inputs. ### Claim 4 (depends on 3) The method of claim 3, wherein preprocessing includes any of the following processes: data cleansing, normalization, rescaling, re-trending, reformatting, and noise filtering. ### Claim 5 (depends on 1) The method of claim 1, further comprising: generating a notification for the determined probability of the forthcoming machine failure, wherein the notification indicates the at least one of: the forthcoming machine failure, and at least one recommendation for avoiding or mitigation the determined forthcoming machine failure. ### Claim 6 (depends on 1) The method of claim 1, wherein the selecting of the machine learning model further comprises: selecting a set of hyperparameters that is optimal for use with the selected machine learning model for optimizing an accuracy level of the machine learning model. ### Claim 7 (depends on 6) The method of claim 6, wherein the selected machine learning model comprises an accurate machine failure prediction capabilities and better long-range predictability of a machine failure with respect to a plurality of machine learning models. ### Claim 8 (depends on 1) The method of claim 1, wherein the determination of the at least a probability is performed on a constant basis. ### Claim 9 (depends on 1) The method of claim 1, wherein the at least one data feature is represented by at least a statistical feature. ### Claim 10 A non-transitory computer readable medium having stored thereon instructions for causing a processing circuitry to perform a process, the process comprising: receiving raw sensory inputs collected from at least one machine; generating a plurality of data features based on the raw sensory inputs; selecting from the plurality of data features a plurality of indicative data features, wherein the selection is based on a distribution of the plurality of indicative data features that determines an association between the plurality of indicative data features and a machine failure, wherein the distribution of the plurality of indicative data features is a set of abnormal parameters of each of the indicative data features; selecting, based on the plurality of indicative data features, a machine learning model; applying the selected machine learning model to the plurality of indicative data features; and determining a probability for a forthcoming machine failure. ### Claim 11 A system for forecasting failures in industrial machines, comprising: a processing circuitry; and a memory, the memory containing instructions that, when executed by the processing circuitry, configure the system to: receiving raw sensory inputs collected from at least one machine; generate a plurality of data features based on the raw sensory inputs; select from the plurality of data features a plurality of indicative data features, wherein the selection is based on a distribution of the plurality of indicative data features that determines an association between the plurality of indicative data features and a machine failure, wherein the distribution of the plurality of indicative data features is a set of abnormal parameters of each of the indicative data features; select, based on the plurality of indicative data features, a machine learning model; apply the selected machine learning model to the plurality of indicative data features; and determine a probability for a forthcoming machine failure. ### Claim 12 (depends on 11) The system of claim 11, wherein each type of the raw sensory inputs is related to at least a process that is associated with the machine. ### Claim 13 (depends on 11) The system of claim 11, wherein the system is further configured to: preprocess the received raw sensory inputs. ### Claim 14 (depends on 13) The system of claim 13, wherein preprocessing includes any of the following processes: data cleansing, normalization, rescaling, re-trending, reformatting, and noise filtering. ### Claim 15 (depends on 11) The system of claim 11, wherein the system is further configured to: generating a notification for the determined probability of the forthcoming machine failure, wherein the notification indicates the at least one of: the forthcoming machine failure, and at least one recommendation for avoiding or mitigation the determined forthcoming machine failure. ### Claim 16 (depends on 11) The system of claim 11, wherein the selecting of the machine learning model further comprises: selecting a set of hyperparameters that is optimal for use with the selected machine learning model for optimizing an accuracy level of the machine learning model. ### Claim 17 (depends on 16) The system of claim 16, wherein the selected machine learning model comprises an accurate machine failure prediction capabilities and better long-range predictability of a machine failure with respect to a plurality of machine learning models. ### Claim 18 (depends on 11) The system of claim 11, wherein the determination of the at least a probability is performed on a constant basis. ### Claim 19 (depends on 11) The system of claim 11, wherein the at least one data feature is represented by at least a statistical feature.