Search plc scada training and you'll notice two things: volumes are healthy (1,800/month for that exact phrase, another 1,700/month for scada training alone), and most of the top results conflate the two disciplines. That's the problem, and it's why most PLC+SCADA courses graduate people who can wire a tag but can't design an alarm strategy.
This guide sorts the two out, maps them to roles, and proposes a 16-week path that produces someone who is genuinely useful on a shop floor — not a certificate-holder with a vague warm feeling about "HMIs."
PLC and SCADA aren't the same skill
A PLC is a small, deterministic computer that reads field I/O and runs a control program every 20–50 ms. It controls things — valves, motors, heaters. It fails safely if you design it properly. The programming languages are IEC 61131-3 — ladder, structured text, function blocks.
SCADA stands for Supervisory Control And Data Acquisition. A SCADA system sits above the PLCs and does the jobs the PLC can't or shouldn't:
- Visualisation for operators (the "HMI" — human-machine interface)
- Historian that logs tag values for days, weeks, years
- Alarming — knowing which of 5,000 tag changes matter, and who to notify
- Reporting — shift reports, OEE, compliance summaries
- Supervisory control — "run recipe B on tank 3 starting at 0400 tomorrow"
A PLC doesn't care what an operator sees. A SCADA system doesn't care about the scan cycle. They collaborate over a fieldbus (Modbus TCP, Ethernet/IP, Profinet, OPC UA) and each does one thing well.
People who learn one and not the other top out at a wall around 18 months into their career. People who learn both stay employable for decades.
Where PLC ends and SCADA begins
This boundary causes more job-interview confusion than any other topic in industrial automation. The short version:
- If it's a 20 ms decision, it belongs in the PLC. Turn the pump off if pressure exceeds setpoint. Close the e-stop loop on a limit switch fault. No operator action required.
- If it's a 2-second decision, the PLC can do it but a SCADA recipe can trigger it. Start the CIP sequence at 0400.
- If it's a 2-minute human decision, it's SCADA all the way. "Operator, acknowledge the high-temperature alarm on Tank 3 before the batch proceeds."
A good PLC programmer knows their code should never rely on SCADA staying online. A good SCADA engineer knows their screens should never depend on sub-second PLC-internal state. The fieldbus is the contract between them, and both sides should treat it as potentially lossy.
The stack, bottom to top
Five layers, in increasing order of latency:
- Field devices — sensors, actuators, VFDs, valves. Physical. Talks DC 0–24 V or 4–20 mA current loops.
- PLC — the deterministic controller. Ladder + ST + occasional SFC.
- Fieldbus — Modbus TCP, Ethernet/IP, Profinet, Profibus. Takes PLC tag values and transports them to whoever asks.
- SCADA / historian — Ignition, Wonderware (now AVEVA), FactoryTalk View, iFIX, OpenSCADA. Logs tags, computes aggregates, raises alarms.
- HMI — operator screens, often running on the SCADA layer itself, sometimes dedicated panels.
Training either discipline in isolation leaves a blind spot. Training both together — with a concrete fieldbus exercise in the middle — produces engineers who understand how data flows end-to-end.
Why SCADA training fails in most courses
Three reasons, from our observation of trainee placement outcomes:
- Courses teach the software, not the design. You get 40 hours of "how to add a tag in Wonderware." What makes an operator screen useful — alarm rationalisation, trend grouping, shift-to-shift consistency — gets a one-hour slide at the end.
- No pairing with a PLC. You drag-and-drop a tag value onto a screen without writing the PLC code that produces that tag. You have no idea whether the value means anything, or what it fails safely to.
- No historian rigour. Trainees finish without ever discussing time-series compression, dead-bands, or why storing every scan-cycle value for 10,000 tags for 10 years will kill your database.
A SCADA engineer who's never written a PLC program is like a front-end developer who's never seen a database. It works until the first real problem.
A 16-week PLC + SCADA study plan
Here's the sequence we'd recommend to anyone starting from scratch. 16 weeks, roughly 8 hours per week.
Weeks 1–8: the PLC half
Work through our complete PLC programming course at half-pace. You'll finish Weeks 1–4 (basics, ladder, timers, sequencing) and stop before PID.
By Week 8 you should have:
- Written ten scenarios from scratch that pass auto-graded tests
- A mental model of the scan cycle that explains every weird behaviour
- A clean mental separation between what belongs in the PLC and what doesn't
Our Basic plan at USD 99/year covers this phase with all 40 scenarios unlocked and the auto-grader active.
Weeks 9–10: fieldbus
Your PLC needs to expose tags to a SCADA system. Modbus TCP is the simplest protocol and still the most common in the field for 2026.
- Lesson: Modbus TCP in 45 minutes — function codes 3, 4, 6, 16 and what they map to
- Exercise: expose four tags from the Motor Start/Stop scenario over a simulated Modbus server
- Exercise: write a Python script that polls those tags every second and prints the values
You'll write the Modbus server as a tiny Node.js sidecar that listens on port 5502 and reads from the simulator's state API. The code is about 30 lines and we'll publish it in a follow-up post.
Weeks 11–12: your first dashboard
Install Ignition Maker Edition (free for non-commercial use) on a Raspberry Pi or a laptop VM. Point it at your Modbus server. Drag your first gauge onto a screen.
- Lesson: Ignition tag browser and tag providers
- Lesson: Templates and UDTs (User-Defined Types)
- Exercise: build a one-screen dashboard for a bottling-line scenario with pump, valve, and tank-level indicators
At the end of this block you'll have a live dashboard showing state from the simulator's PLC program on one side of the screen and an Ignition HMI on the other. The data flows: browser-simulator → its API → Modbus server → Ignition tag → Ignition screen. You've built a real end-to-end control system in software.
Weeks 13–14: alarms and historians
A screen is not a SCADA system until it has alarms and a historian.
- Lesson: Alarm rationalisation — the three questions to ask about every alarm
- Lesson: Dead-bands, compression, and why storing every value is wrong
- Exercise: add a high-high level alarm to the bottling-line dashboard. Make it acknowledge-required.
- Exercise: wire the tank-level tag to the historian, generate a 15-minute trend chart
This is where most trainees discover that alarms are a design skill, not a software feature. You'll tune the alarm thresholds, set the dead-band, add the priority, and argue with yourself about whether to also flash the screen border.
Weeks 15–16: end-to-end
The final two weeks pull everything together into a shift report.
- Build a second HMI screen: "yesterday" — shows a trend of the last 24 hours' tank levels with alarms overlaid
- Write an Ignition report that emails itself every morning at 0600 with the previous day's events
- Record a five-minute screen-capture walk-through of the whole system
Upload the walk-through to YouTube (unlisted if you prefer), paste the link into your CV. Hiring managers will watch 60 seconds of it and be sold.
Which SCADA to learn first
Four realistic choices in 2026:
- Ignition (Inductive Automation) — our recommendation for first-time learners. Free Maker edition for learning, excellent documentation, free certification track. Growing market share.
- Wonderware / AVEVA InTouch / System Platform — still the market leader in process industries. No free tier worth speaking of. Learn it on the job.
- FactoryTalk View (Rockwell) — ubiquitous wherever Rockwell PLCs are. Tight integration with Studio 5000. Not a great free-learning option.
- OpenSCADA / ScadaBR — open-source, useful for hobbyist and academic use, limited market recognition.
For self-study, Ignition Maker is the clear winner. For a career targeted at a specific shop, match the SCADA to whatever the shop runs.
FAQ
Is PLC or SCADA easier to learn?
PLC is harder to learn well but has a smaller surface area. SCADA has a huge surface area but most of it is configuration, not programming. If you learn them in sequence (PLC first), SCADA feels easier. If you try SCADA first, you'll hit a wall on tag semantics within two weeks.
Is there a free PLC and SCADA training online?
Yes. For the PLC half, one guided program runs without an account; a free account has 27 source-tagged practice records and the first six core lessons per dialect, subject to entitlement and rollout. For the SCADA half, Ignition's University of Ignition is free and excellent. Combine them with the 16-week plan above.
What's the best PLC-SCADA certification?
Inductive Automation's Ignition certifications are genuinely respected and free. ISA CAP covers both PLC and SCADA as part of broader automation. Vendor-specific SCADA certifications (Wonderware, FactoryTalk) are costly but gate specific jobs in their ecosystems. See our full PLC certification guide for the broader picture.
Can I get a SCADA job without PLC skills?
You can in the first five years of your career, but you'll plateau. The better SCADA engineers we've worked with can all read ladder logic well enough to tell when the problem is in the PLC vs their HMI.
How long does PLC and SCADA training take?
16 weeks at 8 hours per week if you do both. 8 weeks if you do PLC only. 8–12 weeks if you do SCADA only (and you'll regret skipping the PLC half eventually).
Where to start
- Sign up free for the simulator — 20 minutes in Traffic Light tells you if ladder thinking clicks.
- Follow our complete PLC course for Weeks 1–8.
- Install Ignition Maker. It takes 15 minutes.
- Come back to the Modbus exercise when you've finished Week 8.
Sixteen weeks, two disciplines, one employable skill set. Most of it is free; the fastest-paying part is USD 99/year. There isn't a better ROI in industrial automation training in 2026.
