A diagnostic method includes: acquiring first measurement data based on a physical quantity generated by an object repeating a predetermined operation pattern in a first period; a reference data generation step of generating reference data based on the first measurement data; acquiring second measurement data based on a physical quantity generated by the object repeating the predetermined operation pattern in a second period; and a diagnosis step of diagnosing a state of the object based on the reference data and the second measurement data, in which the reference data generation step includes extracting, from the first measurement data, a plurality of pieces of first period unit data corresponding to at least a part of the predetermined operation pattern, and generating the reference data by calculating a representative value of the plurality of pieces of first period unit data subjected to synchronization processing.

Methods and systems for detecting an abnormal start of a gas turbine engine are described. Speed data points are sampled from a sensor associated with the engine in accordance with a sampling rate, the speed data points being indicative of a rotational speed of a gas generator of the engine during engine start. The speed data points are continuously stored during the engine start. Previously-obtained speed data points which are older than an abnormal start delay are discarded. An abnormal engine start event is detected by comparing a first one of the stored speed data points with a second one of the stored speed data points, the second one of the stored speed data points obtained before the first one.

Aspects of the disclosure relate to computing platforms that utilize improved techniques for dynamic event verification. A computing platform may receive first source data comprising driving data associated with a vehicle over a time period. Based on the first source data, the computing device may determine that the vehicle experienced an event, resulting in an event output. In response to determining the event output, the computing device may generate a request for second source data associated with the vehicle over the time period. The computing device may receive, from a sensor device, the second source data. Based on a comparison of the first source data to the second source data, the computing platform may determine an event comparison output. The computing platform may determine that the event comparison output exceeds a predetermined comparison threshold, and may send an indication of an event in response.

A computer-implemented method for automating the development of industrial machine learning applications includes one or more sub-methods that, depending on the industrial machine learning problem, may be executed iteratively. These sub-methods include at least one of a method to automate the data cleaning in training and later application of machine learning models, a method to label time series (in particular signal data) with help of other timestamp records, feature engineering with the help of process mining, and automated hyper-parameter tuning for data segmentation and classification.

A system including at least one computer and code executable thereby for implementing a mechanism for monitoring performances of applications of an application chain. The system includes an arrangement forming a measuring repository on the one hand for measuring levels of use of resources of applications during periods of degradation of performances of the applications, and by application and by period of the application chain, in a memory storing these levels of use. The arrangement is further operable to: establish a repository of use data by defining and storing in at least one memory, by resource and by application, thresholds of acceptable performance of the level of use of the measuring repository; constitute a categorization module of performance problems as a function of measuring and use repositories; and implement an alert mechanism when the monitoring mechanism detects a performance problem of the applications or when the problem is resolved.

Methods and systems for monitoring the performance of a building management system by analyzing log files of various modules of the building management system. A user request is received at a first one of the plurality of modules of the building management system, which initiates a sequence of messages processed by two or more modules. Each of the sequence of messages include a common tag value that corresponds to the user request. Each of the plurality of modules that process one of the sequence of messages logs the corresponding message including the common tag value in a corresponding log entry. The log entries are analyzed to identify resource utilization of at least some of the plurality of modules.

An artificial intelligence diagnosis system includes a diagnosis model responsive to data received from a plurality of sensors. Each of the sensors is a part of an input channel further including a converter operative to process the received sensor data. A system manager is provided and is operative with an allocator to selectively distribute sensor data to the converters. The system manager operates with the converters such that a converter which is allocated with the sensor data processes the allocated sensor data and inputs the processed sensor data into the artificial intelligence diagnosis model. A converter which is not allocated with the sensor data generates virtual sensor data according to an instruction of the allocator and inputs the virtual sensor data into the artificial intelligence diagnosis model.

The disclosed embodiments relate to a system that automatically adapts a prognostic-surveillance system to account for aging phenomena in a monitored system. During operation, the prognostic-surveillance system is operated in a surveillance mode, wherein a trained inferential model is used to analyze time-series signals from the monitored system to detect incipient anomalies. During the surveillance mode, the system periodically calculates a reward/cost metric associated with updating the trained inferential model. When the reward/cost metric exceeds a threshold, the system swaps the trained inferential model with an updated inferential model, which is trained to account for aging phenomena in the monitored system.

An abnormal irregularity cause identifying device includes a process data acquisition unit that reads process data output by sensors included in a production facility performing a batch stage and a continuous stage, a preprocessing unit that associates a range of a complete timing of the batch stage with an output timing of process data of the process data in the continuous stage based on a residence time of the processing target in the production facility, an abnormality determination unit that calculates an abnormality degree by using process data in the batch stage and process data in the continuous stage associated with each other by the preprocessing unit, and a cause diagnosis unit that determines, for each of the process data output by the corresponding one of the plurality of sensors, whether the abnormality degree calculated by the abnormality determination unit satisfies a predetermined criterion.

The invention relates to a processing, i.e. storing and analyzing, of measurement data. In order to store the measurement data, respective data units or data sets are created for sampling points, a plurality of measurement data temporally adjacent to the sampling-point measurement datum being stored within the data unit. A time interval between time values of the sampling-point measurement data of two consecutive data units at least is set approximately to a multiple of a sampling time interval, i.e. of a reciprocal of the measurement-value recording rate or sampling rate. The method according to the invention reduces the provision of data units because an individual data unit contains, in addition to a sampling-point measurement date, further measurement data temporally adjacent to the sampling-point measurement date. This measure reduces the management of data sets that is required in a database approximately by a factor which corresponds to the number of measurement data within the data unit designed according to the invention.