Bentley Systems, Incorporated

In example embodiments, a hybrid classification/anomaly machine learning architecture is provided that combines a classification model and an anomaly model to perform an engineering task. The classification model and the anomaly model may be used in parallel, and their inference results compared, with their consistency used to improve confidence, and their inconsistency used to detect when additional training or other improvement is required and to capture data useful in such additional training/improvement.

In example embodiments, techniques are provided for using machine learning to predict railroad track geometry exceedances to enable proactive maintenance. A machine learning model of a rail operational analytics application may be trained to directly output a probability of future railroad track geometry exceedances for each portion of track of a railroad. Training may be performed using all available railroad track data, and the task of selecting which data is relevant to predicting probability of railroad track geometry exceedances may be devolved to the machine learning model. Further, assumptions about the specific railroad and data characteristics may be avoided, providing the machine learning model flexibility, and allowing for dynamic changes in the problem formulation.

In various example embodiments, techniques are provided for efficient and reliable anomaly detection and evaluation in a water distribution system (e.g., a smart water distribution system) using both flow and pressure time series data from sensors of the system. The techniques may implement a multi-step workflow that involves decomposing the time series data to remove seasonality and rendering the time series data stationary, detecting outliers of the stationary time series data, classifying sensor events in response to flow or pressure of detected outliers exceeding high or low thresholds for at least a given number of time steps, classifying anomaly events by correlating one or more sensor events related to flow with one or more sensor events related to pressure or by clustering a plurality of sensor events in temporal proximity, and determining a quantitative score for each of the detected anomaly events that indicates a level of significance or importance.

In example embodiments, techniques are provided for efficiently labeling, reviewing and correcting predictions for P&IDs in image-only formats. To label text boxes in the P&ID, the labeling application executes an OCR algorithm to predict a bounding box around, and machine-readable text within, each text box, and displays these predictions in its user interface. The labeling application provides functionality to receive a user confirmation or correction for each predicted bounding box and predicted machine-readable text. To label symbols in the P&ID, the labeling application receives user input to draw bounding boxes around symbols and assign symbols to classes of equipment. Where there are multiple occurrences of specific symbols, the labeling application provides functionality to duplicate and automatically detect and assign bounding boxes and classes. To label connections in the P&ID, the labeling application receives user input to define connection points at corresponding symbols and creates connections between the connection points.

In example embodiments, an enhanced deep belief learning model with an extended Kalman filter (EKF) is used for training and updating a deep belief network (DBN) with new data to produce a DBN model useful in making predictions on a variety of types of datasets, including data captured from infrastructure-attached sensors describing the condition of the infrastructure. The EKF is employed to estimate operation parameters of the DBN and generate the model's output covariance. Further, in example embodiments, the configuration of the DBN model may be optimized by a competent genetic algorithm.

In example embodiments, machine learning techniques are provided for ensuring quality and consistency of the data in a digital representation of infrastructure (e.g., a BIM or digital twin). A machine learning model learns the structure of the digital representation of infrastructure, and then detects and suggests fixes for data errors. The machine learning model may include an embedding generator, an autoencoder, and decoding logic, employing embeddings and metamorphic truth to enable the handling of heterogenous data, with missing and erroneous property values. The machine learning model may be trained in an unsupervised manner from the digital representation of infrastructure itself (e.g., by assuming that a significant portion is correct). An SME review workflow may be provided to correct predictions and inject ground truth to improve performance.

In example embodiments, a software service may employ a neural network to learn a non-linear mapping that transforms element features into embeddings. The neural network may be trained to distribute the embeddings in multi-dimensional embedding space, such that distance between the embeddings is meaningful to the class or category classification, or property prediction, task at hand. The neural network may be trained using weakly supervised machine learning, using weakly labeled infrastructure models. Embeddings for groups may be used to determine prototypes. Elements of an infrastructure model may be classified into classes or categories, or their properties predicted, as the case may be, by finding a nearest prototype.