As presented at ARC Industry Forum and ICC 2026 · Speaking at CODESYS Tech Talk
As presented at ARC Industry Forum and ICC 2026 · Speaking at CODESYS Tech Talk
Signed in as:
filler@godaddy.com
We design control systems so that every value has one origin, one path, and one owner — measured once at the edge, secured at the boundary, and delivered everywhere it is needed without being copied, re-typed, or guessed at. It is how the best plants will run in ten years. We build it today.
The value exists, but it lives in a proprietary driver, a rack no one can touch, or a spreadsheet someone re-types on Friday. By the time it reaches a decision it is a copy of a copy.
The reflex fix for "no visibility" is to add: a bigger PLC, more I/O cards, another cabinet, another server. More hardware, more wiring, more to power, cool, and maintain — and no more information.
Fiber gets cut. Radios fade. Cellular drops. A design that assumes the link is up will lose data, or worse, lose control, exactly when it matters.
Certificates on every field device, renewals nobody tracks, and a VPN that was meant to be temporary. Security that depends on a technician remembering something is not security.
A DCS gets you one source of truth by owning everything: controllers, I/O, historian, HMI and engineering tools from a single vendor, sitting in a single plant. It works when your process is in one building.
Most of ours isn't. A pump station an hour underground, a batch plant with three silos, dozens of stations spread across a state. A DCS answers that with a remote I/O drop over a network it assumes is always up, or a separate telemetry system, which puts you right back to two sources of truth.
We get the same data integrity a different way: open standards at every layer, intelligence at the edge so a station keeps running and keeps its history when the link is down, and hardware your own people can buy, replace and maintain. That is also why it costs a fraction of a DCS to license and to keep, but the price is a consequence of the design, not the reason for it.
Every measurement in a Surety system is taken once, at the device, in its native digital form. Smart transmitters, digitally enabled starters, and IO-Link sensors already carry the process value, device health, diagnostics, and configuration. We do not convert that to 4–20 mA and throw the rest away. We listen.
An edge controller at each station owns that data. It runs the control logic locally, keeps a local history, and publishes upward only what changed — report by exception — so a 48-site network moves a fraction of the traffic a polled system would, and a single-station outage stays a single-station problem.
When the network goes down, and in the field it goes down, the controller keeps running and keeps recording. When the link returns, it forwards what it stored, in order, with timestamps. The site historian never sees a gap. Nobody drives out to reconstruct a shift.
Above the edge, the data moves on open standards — OPC UA and MQTT Sparkplug B — into a single namespace that SCADA, the historian, maintenance, and the enterprise all read from. One name for every tag. One value. One place to fix it if it is wrong.
"Less hardware, more data. That is not a trade-off; it is what happens when you stop converting digital devices to analog and back."

One small panel per station: an edge controller with built-in I/O and networking, smart devices, and a gateway where the devices need one. No rack, no marshalling cabinet, no processor sized for a plant when the job is a sump. In our underground dewatering work this cut hardware and wiring cost per station by roughly a third and made every additional station a copy-and-configure job instead of a design job.
OPC UA, MQTT Sparkplug B, Modbus TCP, IO-Link, IEC 61131-3 code. Every layer in the drawing is replaceable, which is why the same design has moved from a gold mine to a cement batch plant to a gas transmission network without starting over. Integrated and open are not opposites.
Control lives at the edge and does not need the link. History lives at the edge and does not need the link. The link carries changes, not polls, and when it comes back the edge catches the site up. We test this by pulling the cable during commissioning, not by hoping.
Field instruments should not need certificate renewals. We put the security boundary where it can be maintained — at the panel and at the site edge — and make the physical enclosure part of the design. The next section shows how.


Underground dewatering, at a gold mine
Miles of underground pump stations on one small control panel each: groov RIO running CODESYS, smart transmitters through a HART gateway, Ignition Edge for local HMI and history, store-and-forward to site SCADA.
Drives and starters get their own enclosure. The control panel stays small, cool, and quiet.

IO-Link batch plant
Silo levels, valves, pushbuttons, and a signal tower on IO-Link, served as native OPC UA straight to the controller. No translation layer. Safety hardwired and off-network. Eleven spare ports for expansion without redesign.

Dozens of stations, one namespace
Edge nodes at every station publishing by exception over Sparkplug B to an on-premise broker. Three infrastructure decisions made before the first node shipped: where the broker lives, how the master server is sized, and where the boundary sits.

groov RIO and groov EPIC edge controllers. CODESYS runtime, Ignition Edge on board, open protocols out of the box. Surety is an IIoT-Certified OptoPartner.

Ignition, Ignition Edge, and Ignition Edge Panel for HMI, history, and the single namespace above the edge. Surety is a registered Ignition integrator.

Every Surety panel is built on Phoenix Contact power, terminals, and switching. One vendor, one support path, and a bay-sized panel that fits.

IEC 61131-3 control logic, written in structured text and function blocks, portable across the edge hardware we use.
Tell us what you are measuring, where it is, and who needs to see it. We will tell you what we would build and what it would cost to find out.