Back to results
US 11,635,753 B1Granted
Remaining useful life prediction for machine components
AssigneeDIMAAG-AI, Inc.
InventorsRajaram Kudli, Satish Padmanabhan, Fuk Ho Pius Ng
Life of this patent
Estimating the term…
Abstract
Remaining useful life may be estimated for a machine component by training a prediction model, even when limited data from actual failures is available. Feature data such as sensor readings associated with a mechanical process may be collected over time. Such readings may be paired with estimates of remaining useful life, for instance as extracted from unstructured text of maintenance records. Such data may be used to train and test the prediction model.
Claims19
A method comprising:
- receiving via a communication interface a designated time series data set associated with a mechanical process performed by a designated machine including a designated component corresponding with a designated component type, the designated time series data including process outcome data characterizing one or more outcomes of the mechanical process;
- identifying via a processor designated estimated condition information for the designated component at least in part by applying a predetermined ontology to text included in one or more maintenance records associated with the designated machine to produce a plurality of vector representations of the one or more maintenance records, each of the one or more maintenance records associated with a respective point in time;
- determining via the processor an estimated remaining useful life for the designated component by applying a pre-trained prediction model to the designated estimated condition information and the designated time series data including the process outcome data, the pre-trained prediction model being trained to predict historical estimated condition information for a plurality of components of the designated component type using historical time series data associated with a plurality of machines that collectively include the plurality of components;
- and
- storing the estimated remaining useful life on a storage device.
2 more elements
A system comprising:
- a communication interface configured to receive a designated time series data set associated with a mechanical process performed by a designated machine including a designated component corresponding with a designated component type, the designated time series data including process outcome data characterizing one or more outcomes of the mechanical process;
- a processor configured to determine designated estimated condition information for the designated component at least in part by applying a predetermined ontology to text included in one or more maintenance records associated with the designated machine to produce a plurality of vector representations of the one or more maintenance records, each of the one or more maintenance records associated with a respective point in time, wherein the processor is further configured to determine an estimated remaining useful life for the designated component by applying a pre-trained prediction model to the designated estimated condition information and the designated time series data including the process outcome data, the pre-trained prediction model being trained to predict historical estimated condition information for a plurality of components of the designated component type using historical time series data associated with a plurality of machines that collectively include the plurality of components;
- and
- a storage device configured to store the estimated remaining useful life.
1 more element
One or more non-transitory computer-readable media having instructions stored therein for performing a method, the method comprising:
- receiving via a communication interface a designated time series data set associated with a mechanical process performed by a designated machine including a designated component corresponding with a designated component type, the designated time series data including process outcome data characterizing one or more outcomes of the mechanical process;
- identifying via a processor designated estimated condition information for the designated component at least in part by applying a predetermined ontology to text included in one or more maintenance records associated with the designated machine to produce a plurality of vector representations of the one or more maintenance records, each of the one or more maintenance records associated with a respective point in time;
- determining via the processor an estimated remaining useful life for the designated component by applying a pre-trained prediction model to the designated estimated condition information and the designated time series data including the process outcome data, the pre-trained prediction model being trained to predict historical estimated condition information for a plurality of components of the designated component type using historical time series data associated with a plurality of machines that collectively include the plurality of components;
- and
- storing the estimated remaining useful life on a storage device.
2 more elements
Dependencies are inferred from the claim text.