A prediction model generator of a prediction system determines as an explanatory variable, one or more status values among status values associated with a training sample to be used for generation of a prediction model, based on an importance with respect to the training sample, determines an interval to be used for prediction by evaluating accuracy of prediction with the determined explanatory variable with an interval included in a search interval being successively varied, and determines a model parameter for defining the prediction model by evaluating plural indicators for the prediction model defined by each model parameter, with the model parameter defining the prediction model being successively varied, under a condition of the determined explanatory variable and the determined interval.

A prediction server includes: a data save unit that acquires diagnosis data from a diagnosis server; a training unit that constructs a prediction model that predicts a possibility of occurrence of the disorder in an arbitrary area and on an arbitrary date by training a learning model about a correlation among an area indicated by first position information, a date indicated by a diagnosis date, and a type of disorder in a machine learning manner, in which the type of the disorder is a diagnosis result estimated by the diagnosis server based on a diseased portion image by using a diagnosis model trained about a correlation between the diseased portion image and the disorder in the machine learning manner.

A time series data processing system according to the present invention includes a learning unit configured to learn so as to generate a model that takes, of time series data measured from a measurement target, boundary period time series data that is time series data of a boundary period between a normal period and an anomalous period as an input and outputs a teaching signal determined by a preset function in accordance with change of time of the boundary period time series data. The normal period is a period in which the measurement target is determined to be in a normal state. The anomalous period is a period in which the measurement target is determined to be in an anomalous state.

A computer-implemented method of maintaining a transport system in a manufacturing facility including: measuring one or more pose-related properties of a plurality of transportation devices at a plurality of manufacturing stations in the manufacturing facility, at which manufacturing stations one or more workpieces are processed; determining a statistical characteristic for each of the plurality of transportation devices based on the pose-related properties; determining a sequence for maintaining one or more transportation devices based on the statistical characteristics; and optionally performing one or more maintenance procedures based on the sequence.

An information processing apparatus includes an acquisition unit configured to acquire a plurality of time-series data indicating changes in output values of a plurality of sensors from a monitoring target including the plurality of sensors, a model generation unit configured to generate one model indicating a relationship between the plurality of time-series data from each of a plurality of periods in the plurality of time-series data, thereby generating a plurality of models respectively corresponding to the plurality of periods, and a detection unit configured to detect an abnormality of the monitoring target based on the plurality of models and the plurality of time-series data. The plurality of periods can include two periods partially overlapping each other.

A method, apparatus, and system are described. The method includes generating a set of current values associated with at least one component included on a moving vehicle and providing the set of current values over a wireless network. The values are generated by one or more sensors. An edge computing device receives the current values. The method further includes processing the set of current values in real-time using at least one machine learning algorithm to identify a value of a point in time for a failure of one of the at least one component based on the set of current values and at least one set of past values received. The past values are stored in a memory. The set of current values are transmitted with a low time latency between the generating the set of current values and the processing of the set of current values.

The systems and methods described herein provide for a novel deep learning approach to estimating and predicting faulty mechanical system conditions before they occur without using any measurements from the system itself. Environmental data, such as temperature, humidity, occupancy, volatile organic compounds (VOC), equivalent carbon dioxide (eCO2) and particulate matter may be used in the estimation and prediction of faults, failures, and other inefficiencies within the HVAC system.

Techniques for geometric aging data reduction for machine learning applications are disclosed. In some embodiments, an artificial-intelligence powered system receives a first time-series dataset that tracks at least one metric value over time. The system then generates a second time-series dataset that includes a reduced version of a first portion of the time-series dataset and a non-reduced version of a second portion of the time-series dataset. The second portion of the time-series dataset may include metric values that are more recent than the first portion of the time-series dataset. The system further trains a machine learning model using the second time-series dataset that includes the reduced version of the first portion of the time-series dataset and the non-reduced version of the second portion of the time-series dataset. The trained model may be applied to reduced and/or non-reduced data to detect multivariate anomalies and/or provide other analytic insights.

A fault prediction device capable of predicting an accurate deterioration state is provided. A fault prediction device for predicting fault of a target device whose deterioration state transitions with elapse of time includes autoencoders AED1 to AED4 respectively corresponding to deterioration states of the target device. The autoencoder AED2 corresponding to a first deterioration state determines whether the target device exists in the first deterioration state or not based on a state signal indicating a state of the target device. In a case where it is determined that the target device does not exist in the first deterioration state, the autoencoder AED3 corresponding to a second deterioration state determines whether the target device exists in the second deterioration state or not based on the state signal.

An event waveform extracting unit (3) extracts an event waveform from time-series data. A co-occurrence rate calculating unit (4) calculates co-occurrence rates of event waveforms among the time-series data. A grouping unit (5) classifies the time-series data into groups depending the co-occurrence rates of the event waveforms. An event information generating unit (6) determines the time at which the periods during which event waveforms occur overlap with each other among the time-series data included in each group, and generates event information identifying an event related to the event waveforms on the basis of the determined time.