Two Starting Points, One Architecture
Every first-mile data plane deployment starts from one of two positions: a plant that has run for decades on layers of accumulated SCADA, PLCs, and point-to-point integrations, or a new build with no legacy to work around. Brownfield and greenfield are not different products - they are the same architecture applied to different starting conditions, and treating them as identical rollouts is where most deployments run into friction.
Brownfield: Alongside, Not Instead Of
A brownfield plant did not wait for a unified data strategy to start running. Legacy SCADA systems, PLCs on Modbus or S7, and point solutions that have quietly kept the line running for years are not going away on day one - and a first-mile data plane that requires ripping any of it out before it can add value has already lost the argument.
The data plane has to read alongside what is already there: acquiring from existing PLCs and SCADA systems through their native protocols, without requiring configuration changes to systems that are working. It adds structure and a namespace on top of what exists rather than demanding a replacement of what exists. The legacy system keeps running the line. The data plane makes what it produces usable elsewhere.
Incremental Rollout Without Rip-and-Replace
Brownfield adoption succeeds when it can start with one line, one cell, or one protocol and expand from there - not when it requires a plant-wide cutover before anyone sees value. A namespace hierarchy can be defined for the assets being onboarded first, with the rest of the plant added incrementally as connections are configured, rather than requiring the entire ISA-95 model to be complete before ingestion starts.
This incremental path matters operationally as much as technically: it lets a team validate that structure and quality look right on a handful of assets before extending the same pattern plant-wide, and it means a legacy system that has not yet been onboarded is a gap to fill, not a blocker to the systems already running.
Greenfield: Namespace Standardization From Day One
A greenfield facility carries none of that legacy weight - but it carries a different risk: without deliberate namespace design from the first commissioned asset, a new plant can accumulate the same inconsistency a brownfield plant spent years developing, just faster. Every new line, cell, or piece of equipment is an opportunity to either extend a consistent namespace or start a new convention that will need reconciling later.
Greenfield deployments get to make the namespace hierarchy - Enterprise, Site, Area, Line, Cell - the first design decision, not a retrofit. Protocol-native ingestion, automatic topic generation, and consistent schema apply from the first tag onward, so the plant never accumulates the site-specific patches and naming drift that brownfield deployments have to unwind.
What Both Paths Share
Despite starting from opposite conditions, brownfield and greenfield deployments converge on the same properties once the data plane is in place:
- A single ISA-95-aligned namespace that every asset - legacy or new - publishes into
- Protocol-native acquisition that preserves source fidelity, whether the source is decades old or newly commissioned
- Structure and context applied once, at the point of acquisition, rather than reconstructed downstream
- Fleet management that keeps configuration consistent as the deployment grows, one site or one line at a time
What Changes
Brownfield and greenfield are not a choice between two products - they are two entry points into the same architecture:
- Brownfield plants gain structured, namespaced data without disrupting the legacy systems already running the line
- Greenfield plants avoid accumulating the inconsistency that legacy plants spent years developing
- Both scale through the same incremental rollout and fleet management, rather than requiring a single high-risk cutover
- The namespace - not the starting condition - is the long-term architecture either path converges on
Whether a deployment starts with decades of legacy equipment or a blank namespace, the goal is the same: one consistent structure that every asset, old or new, publishes into.