One of the most common applications for real-time data in manufacturing and process industries is SCADA, supervising remote processes over a network. With the growing popularity of cloud computing, many engineers and managers in the automation sector are looking at the possibility of using the cloud for SCADA. Are they being realistic, or just dreaming in technicolor? Is it possible that SCADA will somehow evolve to the cloud?
The acronym “SCADA” stands for Supervisory Control And Data Acquisition. SCADA systems connect sensors and devices in the field or factory floor to an HMI (human-machine interface), allowing plant operators and engineers to view the data in their industrial processes in real time. This interface often supports a supervisory level of coordination and control, such as uploading new recipes to a candy-making machine, changing global settings on a wind turbine, or acknowledging a high pressure alarm for a boiler.
SCADA systems have evolved over time. The first generation systems were “monolithic”, running on mainframe computers, connecting to field devices over proprietary wide-area networks (WANs). The second generation did “distributed” processing, using mini computers communicating with each other over a local-area network (LAN). Communication to the field was still by proprietary protocols on WANs. The current, “networked”, generation uses PCs and open standards such as TCP/IP and open protocols for wide-area networking. Thus it is now possible to access SCADA systems and data from the Internet.
Do you see where this is going? Since SCADA systems have followed the progress of computing in general, and as many view cloud computing as the next logical step in this evolution, enthusiastic visionaries foresee a fourth, “cloud” generation of SCADA, where an entire control system would be running in the cloud.
Back here on earth, most industry experts agree it would be foolish to put the primary control of a power plant, water treatment system, or railyard switching system on the cloud, as it is right now. These kinds of mission-critical control tasks require rugged, reliable data networks and extremely fast response times. Advocates of cloud computing may hope that Internet speed and reliability will eventually support this level of SCADA, but we have no guarantees of that today.
That said, there are other ways of using SCADA, and other uses of process data that lend themselves to real-time cloud applications. Designed properly, with the core requirements for real-time cloud systems in mind, it is possible to put live data from SCADA systems on the cloud in a secure, reliable way. Using specially-designed middleware that supports high data rates and low latency on a data-centric infrastructure, perfectly acceptable real-time performance can be achieved for many types of applications.
Cloud computing can be implemented in different ways. As we explained a few months ago, a private cloud option can be implemented on-site to maximize security, or off-site to reduce costs and gain other benefits associated with the cloud. Another possibility is a hybrid cloud, a combination private and public clouds. With the right kind of infrastructure in place, any of these options could support a system to meet the growing demand for providing access to data from a SCADA system to local or remote users, in real time.
Evolution is a gradual process. It takes time, and it goes step by step. A first step in the evolution towards cloud-based SCADA may well be some kind of cloud-enhanced SCADA. We will talk about that in an upcoming blog, but first we need to clear up some of the fear, uncertaintly and doubt surrounding the discussion about SCADA and the cloud.