SCADA is the layer above the controllers: supervisory control systems that collect data from PLCs and RTUs, present it to operators through human-machine interfaces (HMI), and write setpoint changes back down. The craft splits into operator panel graphics, alarm management, historian configuration, telemetry, and the SCADA cybersecurity work of keeping remote access from becoming remote control. MarketsandMarkets values the SCADA market at USD 12.89 billion in 2025, growing at 9.2% annually to USD 20.05 billion by 2030, with utilities, oil and gas, and manufacturing as the main buyers . The people who build and defend these systems are hired on a different question than the market answers: what the operator sees at three in the morning.
Challenges in SCADA Recruiting
Supervisory control systems span substations, pipes, and production lines
The first split in the discipline is architectural. A utility SCADA reaches across a territory: RTUs at remote stations, radio or cellular telemetry, and a control room that supervises hundreds of sites it cannot walk to. A plant SCADA sits on a control LAN a floor above the PLCs, with a historian attached and no telemetry problem at all. A pipeline SCADA is a third thing, stretched along a right-of-way with valve sites polled on a schedule. The same title, three estates. Engineers move between them poorly because the failure modes differ: a utility engineer worries about polling deadbands and lost communications, a plant engineer about screen response and tag counts, and each is hired against a market growing at 9.2% a year that has not produced enough of either . The job description that does not name the estate receives candidates for all three, and half of them will not recognize the site on arrival.
Operator panels carry the alarm burden no server screen shows
Alarm systems are where SCADA engineering earns its money, and most of them are broken in the same way. The alarm management lifecycle defined in ANSI/ISA-18.2 exists because ineffective alarm systems have been documented as contributing factors to major process accidents . The standard's target numbers are unforgiving: no more than one to two alarms per ten minutes for a single operator, fewer than ten in any ten-minute window, and a flood condition less than 1% of the time . Real systems run far over those numbers, and the result is operational: operators stop trusting the horn, shelve alarms they should act on, and miss the one trip that matters. The engineer who can rationalize an alarm database down to the standard is a scarce and specific hire, distinct from whoever builds the graphics.
Touch panel programming separates graphics from control
Touch panel programming and HMI development look like design work but are engineering with a standard attached. ISA-101 frames the human-machine interface as a lifecycle deliverable that spans philosophy, style guide, implementation, operation and maintenance, aimed at detecting abnormal situations early rather than depicting the plant prettily . The consequence for hiring: a graphic designer who cannot read a tag database cannot ship a screen that matters, and a controls programmer who ignores display hierarchy ships screens operators cannot navigate under pressure. Both failures show up the same way, as a control room that works around its own interface. The people who do both well are the ones ISA-101 was written by.
Industrial data acquisition scales with tag counts, not dashboards
Industrial data acquisition has a scale problem hidden inside its name. A dashboard with a hundred points is a weekend's work; a historian holding a hundred thousand tags, with compression, deadbands, event retention and millisecond timestamps surviving a failover, is an engineering discipline. Polling schedules, scan classes, and the difference between a value that is stale and a value that is zero all matter once the plant starts making batch records from the historian. So does interface engineering: which protocols the RTUs speak, how exceptions are reported versus polled, and what happens to data during a network partition. CVs rarely say which end of that scale a candidate lived at, and the interview that does not ask is hiring on faith.
SCADA cybersecurity stops at the RTU in most CVs
The threat record keeps SCADA cybersecurity an active discipline. In its December 2025 joint advisory, CISA and allied agencies describe pro-Russia hacktivist groups scanning for internet-facing VNC ports, brute-forcing default or weak passwords, and reaching live HMIs to change setpoints, disable alarms, and force loss of view; intrusions were recorded as recently as April 2025 against water, food and agriculture, and energy targets . The remediation is not exotic: segmentation, DMZs, MFA, time-limited remote access, default-deny firewalls. The scarce hire is the engineer who can apply those inside a system that must keep polling, because a SCADA network that loses a poll is not secure, it is blind. Security staff who only know office IT add the most common failure of all: a patch window that costs the plant its view.
Industrial displays expose designers who never sat a night shift
Industrial displays carry the plant's situation awareness, and the standard for them is written in grey. High-performance HMI practice replaces schematic art with grey backgrounds, embedded trends, analog indicators showing normal versus abnormal range, and a four-level display hierarchy; tests run with EPRI and the ASM Consortium found operators detected abnormal situations earlier and handled more of them successfully, which the authors translate into hundreds of thousands of dollars per year per plant . A designer who never sat a night shift ships screens that are beautiful at nine in the morning and useless at three. That experience shows in a portfolio in ways a CV cannot hide, if the person reading the portfolio knows what grey means.
SCADA integration evidence hides in migration cutovers, not greenfield builds
Assessment ends where the estate begins: which system was migrated, how the tag database was converted, what happened to historian continuity across the cutover, and which screens broke on the first shift after go-live. A greenfield SCADA built in a lab proves configuration skill; a migration proves industrial monitoring judgement, because it happened on a system operators were using. The probes that separate owners from witnesses are specific: who owned the alarm rationalization, who decided the polling architecture, and what the rollback position was when the first cutover attempt failed. The cost of skipping those questions lands in the control room, where an unrationalized alarm system trains operators to ignore alarms, and the resulting missed trip is paid in downtime or worse. False negatives have their own price: an integrator who has shipped ten clean cutovers is not interchangeable with one who has watched ten.
References
- SCADA Market Size, Share, Trends and Growth Analysis, 2025-2030 — MarketsandMarkets. (accessed 2026-09-28)
- ANSI/ISA-18.2-2016, Management of Alarm Systems for the Process Industries — International Society of Automation (ISA). (accessed 2026-09-28)
- Implement an Effective Alarm Management Program — exida. (accessed 2026-09-28)
- ISA-101 Series of Standards — International Society of Automation (ISA). (accessed 2026-09-28)
- Pro-Russia Hacktivists Conduct Opportunistic Attacks Against US and Global Critical Infrastructure (AA25-343a) — Cybersecurity and Infrastructure Security Agency (CISA). (accessed 2026-09-28)
- The High Performance HMI Handbook Overview — International Society of Automation (ISA). (accessed 2026-09-28)