Plant floor data consists of several thousand data points collected each second. Data normally comes from raw sensors (which have a very high sampling rate), controllers, process variables and assembly machines. Relational databases (RDBs) support simple operator queries, answering questions such as “What is the largest shipment today?” They are built to manage relationships and are ideal for storing contextual or genealogical information, but not the best approach for vast amounts of process data collection and optimization.
Advantages of Historians over RDB-
*Built-in data collection: Manufacturing systems are closed and proprietary. A custom code has to be written for each type of equipment. RDB’s are not designed for built in data collection thus custom. Historians can be compliant with manufacturing standards like Object Linking and Embedding for Process Control (OPC) and MTConnect for near real-time data collection from manufacturing equipments and sensors. Historians are designed for manufacturing and process data acquisition. They maximize the power of time series data and are a perfect solution for answering questions that manufacturing typically needs to address real-time decisions in production.
* Operational excellence: One can easily get answers to questions like power consumed by a particular machine or process delay times in the last week. The raw data can be rolled up to key performance indices like OEE, SPC, Cpk etc… Trending and analysis of historical data allows root cause analysis in case of a failure.
*Higher data compression: Historians collect huge amount of raw sensor data. Better compression technique requires less space and thus reduce cost.
*High availability: Historian allows data redundancy through buffering the data right at the source. This allows high availability and reliable data even in case of network failure.
*Better data security: Historians are designed to address regulatory requirements such as the 21 CFR Part 11 by implementing electronic signatures.
Reference: The Advantages of Plant-wide Historians vs. Relational Databases Comparing Two Approaches for Data Collection and Optimized Process Operations, White Paper, GE Fanuc.