Data has become a critical part of almost every industry. Data drives innovation, provides traceability, and even assists in troubleshooting. As a company grows, so does its data, which is why companies turn to historian solutions. Choosing a historization platform is an important undertaking, and there are a few key factors to keep in mind when picking a data historian.
Data Structure and Storage
There are two types of data that is typically historized: time-series and event-based.
Time-series data is the typical process data from each device on a piece of equipment, collected at specific rates (polling rates). Since time-series data accumulates very quickly, historians employ several methods to efficiently store the data. One method is to limit incoming data by implementing deadbands to reduce noise and repeated data points. The bounds are often tied to the resolution of the transmitter and are typically configured per tag. Another method is to compress archived files to reduce their virtual footprint. Many historians run a loss-less compression algorithm, similar to zipping a file in Windows. The downside is that there is often a trade-off for efficiency when retrieving data for viewing.
Event-based data is select data captured only under specific circumstance (triggers) and is often stored in relational databases such as SQL which are designed to handle millions of entries. Some examples of event-based data are batch reports, audit trails, and alarms.
One important detail often overlooked is how data will get from the equipment to the historian. Two of the most widely used protocols are OPC and MQTT. OPC can be subdivided into OPC DA (Data Access) and OPC UA (Unified Architecture). OPC DA is built on Windows DCOM which is an aging technology, but still widely used. OPC UA offers several advantages over OPC DA including increased security and platform independence. OPC typically uses a polled model in which the OPC server pulls the tag data from the source at defined intervals. MQTT with Sparkplug B is an emerging lightweight protocol that is taking hold in the IIoT sector. MQTT follows the publish/subscription model through a centralized broker, offering simple configuration, increased security, and local buffering. Other types of data collection are also sometimes necessary including querying databases or importing from flat files (ex. csv). When picking a historian, it is important to choose a historian that supports the protocols of the equipment or an intermediate software (middle-ware) is required to translate.
Security and Auditing
Data security is paramount for most companies. Therefore, historians take security very seriously. For most historians, archived time-series data is stored in an encrypted, proprietary format. Event-based data can be stored in a standard relational database, like SQL, or in a proprietary format. Encryption through secure ports may be used for data transfer to and from the historian.
Access to the configuration application and data stores is controlled through security mappings within the application that can be tied to a company’s active directory. The resolution of access restrictions varies between software, but all historians have a baseline of read only, write only, and read/write access.
Visualization and Context
Data is only useful if end users are able to view and analyze it. Most historian solutions offer a visualization option of configurable objects such as charts and graphs. Using those objects as building blocks, screens can vary in complexity, from simple overview dashboards to detailed overlay analysis screens. Since raw data can be very difficult to parse and utilize, historians allow the data to be put into context through logical groupings, association to meta-data, applying calculations, or combining with event data. Through these methods, system administrators can organize the data in meaningful ways for the end user including hierarchical relationships that allow for templatization and comparisons between similar equipment. The historical data can also be served to other applications including SCADA, reporting, and PAT systems through web-API, ODBC, and OPC-HDA.
Historian systems can play a key role in a company’s growth and development. Historians are optimized to support multiple data formats through the industry’s most prominent protocols. They offer a secure repository, utilizing up-to-date encryption methods, and provide powerful analysis tools. Visualization options can render insights into company processes from any corporate level. In addition, the data can be made available to third party systems to allow further capabilities. Choose a historian that will not only store and protect data, but also maximize the utility of that data.
Michael Lehrich is lead engineer at Panacea Technologies, a certified member of the Control System Integrators Association (CSIA). For more information about Panacea, visit its profile on the Industrial Automation Exchange.