Watching a process you cannot see
Some processes are too large or too scattered for anyone to watch directly: a water network with pumps and tanks across a region, a pipeline hundreds of miles long, an electrical substation network, or a big plant with dozens of production areas. SCADA, short for supervisory control and data acquisition, is the kind of system that gathers readings from all those places, shows them to operators in one control room and lets those operators send commands back.
The key word is supervisory. SCADA usually does not run the fast, detailed control of each machine. That job belongs to local controllers. SCADA sits above them, collecting information and allowing oversight and adjustments.
The main parts
A typical SCADA system has several layers.
Field devices. Sensors measure things such as level, pressure, flow, temperature or electrical values. Actuators such as valves, pumps and switches do the physical work.
Local controllers. At each site, a programmable logic controller or a similar remote terminal unit reads the sensors, runs local control logic and holds the data. A remote unit at a distant pump station, for example, might keep a pump going properly even if the connection to the control room drops. Those controllers are often rugged units like those described in the guide to logic controllers.
Communications. A link carries data between sites and the control center. Depending on distance and cost, that may be wired networks, leased lines, radio, cellular IoT or satellite. Spread-out sites often use whichever link is available, so the network can be a mixture of old and new.
The central servers. Software at the control center polls or receives data from the field, stores it and applies alarm rules.
The operator interface. Screens, often called HMI (human-machine interface), display diagrams of the process with live values, colors that change with status, and controls to open valves or start pumps.
A historian. A database that records readings over time, generally as time series data, so trends and past events can be reviewed.
What the system does all day
- Polling and reporting. The central system may ask each site for its latest values at set intervals, or sites may report when something changes. Slow or unreliable links are one reason designs vary.
- Alarms. If a tank level goes too high or a pressure drops out of range, the system alerts operators. Good alarm design matters: too many alarms cause operators to ignore them, which is a well-known problem in control rooms. This is a cousin of the dashboards and alerts issues in any monitoring setup.
- Control. Operators may change set points, start or stop equipment, or acknowledge alarms. Sensitive commands often need confirmation and are logged.
- Records and reporting. Historical data supports audits, billing, planning and investigation after an incident.
Where SCADA is used
It is common in utilities and infrastructure: water treatment and distribution, wastewater, electricity transmission and distribution, oil and gas pipelines, and rail and transit systems. Large factories and processing plants also use SCADA-style supervision, though what they call it varies. Many utility networks rely on this kind of supervision, in the same family as the systems that detect and respond to outages.
How SCADA differs from modern IoT
SCADA predates the term Internet of Things by decades, and the two overlap. Both gather remote readings and act on them. Differences tend to be about culture and design:
- SCADA emphasizes dependability, predictable behavior and long service, and often uses specialized industrial protocols.
- IoT platforms often emphasize cheap devices, flexible cloud services and rapid change.
- Many organizations now blend them, feeding SCADA data into analytics and reporting, or using lighter IoT sensors to extend coverage to places where full SCADA hardware would be too costly. Differences in protocols can make device interoperability a real task.
Security and reliability concerns
Because SCADA runs essential services, mistakes and attacks can have wide consequences. Several issues recur:
- Old protocols and equipment. Some industrial protocols were designed for isolated networks and lack authentication or encryption. Equipment can remain in service for decades, making updates difficult.
- More connections. Linking SCADA to corporate networks or remote-access tools for convenience raises exposure. The difference in priorities between plant and office networks, covered under operational technology, explains why this needs care.
- Common good practices. These typically include network segmentation, limited and monitored remote access, removing default passwords, tracking firmware versions and keeping tested backups. Specific design should involve qualified specialists.
- Operational resilience. Systems are often built with redundant servers and communication paths, and sites can usually keep running locally if the link to the control room fails. Planning for offline operation is part of good design.
Reading the term correctly
People sometimes use SCADA loosely to mean any industrial control or monitoring software. Strictly, it refers to the supervisory layer that collects data from many points and lets operators oversee and command a process. If a system watches one machine and never reaches beyond the cell, it is probably better described as a local control system with an operator screen.
Understanding the layers, from sensor to controller to network to control room, makes the rest of industrial automation easier to follow.