Stop Paying the Integration Tax: What DataReady Actually Does

Building context at the controller

The key to how DataReady works in practice is Smart Objects, and they deserve more than a passing mention. Smart Objects define what data to collect, when to collect it, and how to group it so that structure and context are built into the information before it ever leaves the controller.

Rather than sending every intermediate signal upstream, which is what happens when a single fault ripples across devices, drives and HMI layers generating multiple competing records, Smart Objects focus on publishing the meaningful event with the right context already attached. The difference between a dataset that needs hours of cleaning and one an operations team can work with straight away comes down to decisions made at this level.

Anchoring time where the event happens

Timestamps are part of that same story. When events are time‑stamped at the controller where the data is actually recorded, you get a consistent reference point that carries through the entire information chain. That sounds simple, but anyone who has tried to correlate events across a mixed line knows how quickly things fall apart when the controller, device, HMI and historian each tell a slightly different version of what happened. Anchoring time at the source is a practical improvement that pays off at commissioning and every time something goes wrong later.

The other thing Smart Objects enable is genuine contextualisation at the edge. The goal is not to collect more data, it is to make the data that leaves the controller meaningful without someone having to manually decode it at the other end. When machine information is organised into models that describe what each value represents and how it relates to everything else, it stops being a flat list of tags and starts being something an operations team can act on. That also makes it much easier to scale from one machine to a line or a fleet without reinventing the same integration logic on every project.

Delivering the information customers actually want

The information models that DataReady targets are things customers are already asking for. Asset identification and overview, OEE, maintenance events, and WAGES across water, air, gas, electricity and steam are all established priorities that tie directly to cost, performance and reporting outcomes.

What DataReady adds is a standardised, repeatable way to deliver those models consistently across different machines and different projects, built into both the controller and FactoryTalk Optix, rather than being rebuilt from scratch each time.

What this changes for OEMs and System Integrators

The opportunity for SIs and OEMs to support their customers follows naturally from there. A DataReady implementation means instead of handing over a tag list and a manual, you hand over a machine that publishes consistent, structured information that integrates cleanly into whatever the customer uses downstream. That has real value at commissioning time and every time the customer wants to expand a line, add a new system, or compare performance across a fleet.

For anyone wondering where to start, keep it simple. Pick one machine or one machine type you build regularly, implement Asset ID and one other model, usually OEE or maintenance events, and focus on getting clean contextualised information flowing from the controller. Validate that it works in two places: locally for operators and technicians, and upstream in whatever the customer already uses. Once you can show that the same structured dataset supports both, you have a repeatable pattern worth scaling.

The broader point is that this kind of change lifts the whole ecosystem. Everyone has their own libraries and standards, and that is not going to change overnight. But when machines publish information in a consistent, contextualised way, integration becomes less of a custom exercise every time and more of a predictable, repeatable handover. You do not have to commit your entire machine standard to it. Start with one machine, try one model, and see what it changes. The only way to know if it makes your projects easier is to give it a go.