Historians are the heart of Big Data inside the plant. Without a central repository for data to be queried, the IoT is limited. However, plant historians often use archaic licensing and hosting schemes. Current historian prices for the largest and most robust historian packages are tag-based and expensive. As the IoT increases the number of data points needed to be recorded, the licensing costs will spire out of control. As historians are becoming more ubiquitous we need to encourage vendors to update their licensing schemes, so that small and medium manufacturing facilities can take advantage of large data point counts.
The advances in information gathering that are coming from the advent of the Internet of Things (IoT), offer incredible potential for the modern engineer to see inside manufacturing at a detail level never before possible. However, the IoT is a promise of things that are already being underutilized in current manufacturing. The future of the IoT doesn’t start tomorrow when new, tiny wireless sensors are readily available; it starts today, by improving tools already available and how they are used. I would like to highlight the two key tools essential to the IoT and are ready for improvement today: historians and data analysis packages.
Historians are the heart of Big Data inside the plant. Without a central repository for data to be queried, the IoT is limited. However, plant historians often use archaic licensing and hosting schemes. Current historian prices for the largest and most robust historian packages are tag-based and expensive. As the IoT increases the number of data points needed to be recorded, the licensing costs will spire out of control. As historians are becoming more ubiquitous we need to encourage vendors to update their licensing schemes, so that small and medium manufacturing facilities can take advantage of large data point counts.
Another piece of the historian puzzle is that historians usually require large centralized servers and trained IT administrators to handle the maintenance. These are tools that are most often available to companies that operate at scale, but not to the large number of medium- to small-scale facilities. As IO is moving to a network distributed model, which allows for ease-of-scaling and low-cost implementations, so too, does the modern historian need to match this model. On-machine self-contained historian modules are easy to maintain, can continue to track data during network interruptions, and can be aggregated in a central historian using a multitude of methods that offer flexibility to handle the exponential increase in data that comes with implementation of large quantities of IoT devices. On-machine historians provide many answers for small facilities without large IT departments, as well as large manufactures who need the ability to scale their processes to meet customer demand. These types of historians are still in their infancy relative to the state of modern manufacturing data-gathering strategy and need to begin to be utilized today, as we prepare for the coming of the IoT.
Extending from the need for more robust historian implementations is the need for data analysis on the new information that is being provided by the IoT. Great strides are being made in the industry regarding the realm of data visualization. It is easier than ever to create charts and graphs that quickly and efficiently show data streaming in from the plant floor to the entire employee chain, from engineering, to supply chain managers. This is a fantastic advantage, but does not help to correlate the data points to help us turn the visualizations into actionable items and initiatives that increase plant efficiency.
The effort to perform the correlation analysis and the proper links back to actual causation is still often a herculean effort and only becomes more complicated as the number of data points that need to be sorted, increase with the IoT. To provide aid to this problem, one begins with developing strategies not only for implementation of data analysis software packages, but with advanced process design.
Starting with the initial stages of process design, the modern engineer needs to carefully analyze where to place and implement data-gathering devices to ensure that big data is not further confused with useless data points that bog down correlation and causation efforts. Coupling good, data-gathering strategies with new data analysis packages, is a necessary skill and can be implemented today even before the commencement scale-up of IoT.
The Internet of Things (IoT), offers exciting advantages for the future with it's ability to increase data acquisition, but at the heart of it, it is only an extension of technology that already exists today. Refining and utilizing existing historians and data analysis can bring existing manufacturing technologies to their full potential, especially since the increase of data has the potential to greatly increase GDP and manufacturing efficiency. However, if we wait for the IoT to come to us and we do not respond today, using the tools we already have, than the process will only become more complicated and difficult as large piles of unsorted data grow and outpace our ability to use it for a brighter future.
Hide