上QQ阅读APP看书,第一时间看更新
Disadvantages of Historians
Historians were developed to collect data coming from the whole plant and often from all plants of the same company. In general, they are designed to manage and make available a large amount of data. This means that optimizations are made with respect to data sampling and storage. This might produce the following disadvantages:
- The data collected by the Historians might be a subset of the data collected by SCADA or the different PLC concentrators; sometimes, not all the data that is available with other data-gathering techniques is collected and made available by Historians.
- Data is stored using filtering and compression mechanisms. This implies that when we get data from Historians, it is stored as raw data, which does not provide a snapshot of an asset at a specific time as an outcome. This is something that we must deal with, since typically the data-stream analytics running in the cloud work with snapshots of data that are temporally consistent and related to an asset.
- The data-collection configuration of Historians usually implements a lower sampling rate than SCADA systems to minimize the bandwidth consumption and optimize the data storage. This means you might have a data collection time of a minute or even more. This is something that you have to take into consideration when developing analytics on the cloud.
- Connecting directly to a Historian system implies that we can talk with it according to its API or SDK. It has the same disadvantages for developing and maintaining the related connectors that we already analyzed for the PLC and SCADA. There are some Historians that are very common in the market, such as the PI System of the OSIsoft. Another option for using the Historian as a data source for the I-IoT data flow is to connect to it through its OPC server. In this case, if we choose to use OPC Classic, we must use OPC Historical Data Access (HDA) Server. This was born to exchange archived process data rather than real-time data such as the OPC DA, so you can query for batches of aggregate data on a regular basis. The same considerations and suggestions that we explained in the previous section about how to deal with the OPC are also valid for the Historians.