False Non-Arrival
Ambyint’s Platform is designed to flag anomalies in well behavior and operating conditions enabling users to identify deviations which could result in incorrect optimization responses, production lost/deferred, or a more costly operational issue. The anomaly detection components of Ambyint’s InfinityPL leverages artificial intelligence (AI) and machine learning (ML) to recognize patterns in the data much like any production engineer would while performing recurring well surveillance.
False Non-Arrival - The False Non-Arrival anomaly was developed in response to repeated customer feedback and interactions, where users (both Ambyint and customer) manually identified cases in which a non-arrival (i.e. no arrival detected by the arrival sensor) appeared inaccurate. These instances could be contradicted by recognizable patterns, inflection points, in tubing pressure and rate data. With numerous examples of both true non-arrivals and false non-arrivals available in the Ambyint platform, Ambyint used supervised machine learning to label these events and train a model to automatically detect false non-arrivals with great accuracy.
Example of false non-arrival
Example of true non-arrival
Want to learn more about our InfinityPL anomaly detection? Let’s talk - get in touch with us today for more information.
