China
Africa
Explore chapters and articles related to this topic
Operational and Business Analytics Integration
Published in Osvaldo A. Bascur, Jim O’Rourke, Digital Transformation for the Process Industries, 2020
Osvaldo A. Bascur, Jim O’Rourke
In process plants and refineries, time-series databases are most prevalent for operations, maintenance, and engineering use cases, with supporting relational databases. Time-series data is used for finding root causes of plant problems, production reporting, calculating key performance indicators (KPIs), and so on. Conversely, business databases are more transactional, more interrelated, and house data such as production costs. If the enterprise wishes to more clearly understand and make impactful decisions regarding total production costs, time-series operations data is combined with business data for greater insights.
AI and ML applications in the upstream sector of the oil and gas industry
Published in Manan Shah, Ameya Kshirsagar, Jainam Panchal, Applications of Artificial Intelligence (AI) and Machine Learning (ML) in the Petroleum Industry, 2023
Manan Shah, Ameya Kshirsagar, Jainam Panchal
Storing massive quantities of data collected by a shale drilling site was formerly prohibitively expensive (Tahmasebi et al., 2017). Both on-premises and cloud storage costs have dropped significantly. On-premises storage is usually located on a server-class PC that is hooked to the monitoring PC (Syed et al., 2021). One of several common time-series databases, such as OSIsoft Pi, is installed on the server-class PC. Time-series databases, unlike relational databases, are capable of efficiently storing large volumes of real-time data.
BIM- and IoT-based monitoring framework for building performance management
Published in Journal of Structural Integrity and Maintenance, 2018
Kai Kang, Jiarui Lin, Jianping Zhang
The existing databases can be divided into two categories: relational databases and NoSQL databases. The traditional relational database is based on the relational model, using structured query language (SQL) as the query language, and suitable for storing highly structured data. Over the past decade, with the development of the information technology and industry, especially the rise of the mobile Internet, there are more complicated data types, more diversified storage requirements, and substantially increasing data volumes. The relational databases gradually show their disadvantages and become difficult to meet these new storage needs. Therefore, NoSQL databases get into a period of rapid development. Many database systems are designed for different kinds of data, and the form of data organization becomes more flexible. According to the data type and manner in data storage, NoSQL databases can be divided into the following categories: (1) document-oriented database, which organizes the data into documents without strict requirement to the structure, such as MongoDB and CouchDB, (2) key-value database, which uses key-value model and is suitable for many scenario depend on its implementation, such as Redis, (3) column oriented database, which stores data by column to get higher performance in queries by column, such as Cassandra, (4) graph database, which is designed for graph data such as Neo4j, (5) others, which contains many databases with different design targets. For example, InfluxDB and OpenTSDB are time-series databases, and these types of databases are designed for time-series data to achieve unique advantages over common databases (InfluxData, 2017).