As organizations continue on their digitalization journey, there are more and more words used to describe the environment, the technology, and the process. Here’s a dictionary to help you on your journey.
Machine learning is an analytical method that is a part of artificial intelligence. Systems are able to learn from data using different techniques to identify patterns and provide information for decision-making or action. Machine learning is divided into different types including unsupervised machine learning, supervised machine learning, and reinforcement learning.
Unsupervised learning is a type of machine learning. It is characterized by the fact that we don’t have pre-knowledge of the output, only the input. HanPHI uses unsupervised learning by taking your historical normal, fault-free data to create patterns and expected values.
Modeling is the creation of a representation of something that can be used for monitoring, simulation, forecasting, and other needs. There are different types of modeling including empirical, neural, physical, statistical, and structural modeling.
HOW HAVE YOU IMPLEMENTED AI ELEMENTS AT YOUR ORGANIZATION?
Unstructured data is information that does not fall under a prescribed format. An example of this are images or audio files. You can categorize them by their type (file format) but the content includes qualitative data.
Semi-structured data is a combination of both structured data and unstructured data. A good example is an email–it includes structured data like the time the email was sent but also unstructured data like the body of the message.
A communication protocol is a set of standards or rules that enables two systems to talk with each other (ie send information). Depending on the systems that need to talk with each other, there are various communication protocols used. Some examples are OPC, Modbus, EtherNet/IP, and BACnet.
A data historian collects real-time data from equipment and archives the data into the database. The top data historians provide historical and real-time data without loss, built-in data visualizations, data collaboration tools, automated reporting, and APIs.
APM stands for asset performance management. The goal of APM is to improve reliability and availability of assets. There are different types of APM solutions including condition-based monitoring, predictive maintenance, and reliability-centered maintenance. For HanAra, our APM solution HanPHI identifies potential and hidden failures by creating an equipment and plant health index.
FDD stands for fault detection and diagnostics. The goal of FDD is to minimize risk and improve operations through monitoring, fault detection, and diagnostics. FDD can be divided into the following stages: 1) monitoring, 2) fault detection, 3) fault identification, 4) fault diagnostics, and 5) evaluation and analysis. For more details, check out our blog post on fault detection and diagnostics.