Quick summary
Most industrial IoT projects do not start with a blank factory floor. They start with decades of installed equipment that was never designed to connect, and the real engineering is bridging that gap safely. Edge computing and protocol translation make it possible without ripping out working machinery.
The promise of industrial IoT is compelling: real-time visibility, predictive maintenance and data-driven optimisation across the production line. The obstacle is that the machines already on the floor were built for reliability and longevity, not connectivity, and they will remain in service for years.
This is the brownfield problem. Connecting a fleet of legacy controllers, sensors and drives to a modern data platform is rarely about installing something new on a clean slate. It is about meeting old equipment where it is, translating what it can say into something a modern system can use, and doing so without compromising the operation or its security.
The economic case for getting this right is substantial. McKinsey estimates that the Internet of Things could generate up to 12.5 trillion dollars in value globally by 2030, with factories alone accounting for up to 3.3 trillion dollars, just over a quarter of the total (McKinsey, 2024).
Yet that value has proved stubbornly hard to capture, particularly in industrial settings. The reason is rarely the technology itself. McKinsey's analysis points to interoperability and cybersecurity, not raw capability, as the factors that decide whether projects move from pilot to scale. The implication is that the integration layer, the part that connects old and new, is where industrial IoT succeeds or stalls.
The hardest part of industrial IoT is not the cloud platform or the analytics; it is the connection to equipment that predates the entire concept.
This is why brownfield integration is a discipline in its own right. A greenfield deployment can specify connected equipment from the start, but most manufacturers and operators are working with what they already own.
Takeaway: The largest share of industrial IoT value sits in factories, but capturing it depends on interoperability and security at the integration layer, not on the analytics platform itself.
Legacy industrial equipment resists connection for three intertwined reasons. First, it speaks older or proprietary protocols, from serial fieldbuses to early industrial standards, that modern systems do not natively understand. Second, it was designed with no security model, because isolation was assumed. Third, its operational lifecycle is measured in decades, so replacement is neither quick nor cheap.
These constraints rule out the obvious approach of simply swapping everything for connected equipment. A production line cannot be halted for a wholesale rebuild, and the capital cost of replacing functioning machinery rarely justifies itself on connectivity grounds alone.
The practical consequence is that integration must be additive. Rather than replacing the controller that runs a press reliably, the goal is to extract its data and expose it to modern systems while leaving its core function untouched. This is where a managed approach to connecting field devices into a data platform earns its place, because the value comes from the data the equipment already produces, not from replacing the equipment that produces it.
Takeaway: Legacy equipment resists connection through old protocols, absent security and long lifecycles, so integration must add a data layer rather than replace working machinery.
Successful brownfield integration tends to rely on a small set of proven patterns, applied according to what the equipment supports:
Protocol gateways that translate legacy fieldbus or serial data into modern formats
Edge computing that processes and filters data close to the machine before sending it onward
Non-intrusive data capture that reads from equipment without altering its control logic
Local buffering so connectivity interruptions do not lose data or disrupt operations
Edge computing deserves particular attention. By processing data at the machine rather than sending everything to the cloud, it reduces latency, limits bandwidth costs and keeps sensitive operational data closer to home. The reason this matters is that many industrial decisions need to happen in milliseconds, faster than a round trip to a remote server allows.
The art lies in matching the pattern to the asset. A modern drive may expose a clean data interface, while a decades-old controller may need a gateway that speaks its dialect, and a well-designed integration accommodates both within one architecture.
Takeaway: Protocol gateways, edge computing and non-intrusive data capture are the core patterns, and good integration matches each pattern to what the specific asset can support.
Connecting previously isolated equipment to a network removes the protection that isolation once provided. Every newly connected device is a potential entry point, and in an industrial context the consequence of a breach can be physical, not merely financial.
The standards community has recognised this directly. In 2025 the IEC published dedicated guidance, IEC PAS 62443-1-6, on applying the 62443 security series to the Industrial Internet of Things, a signal that connecting industrial equipment to IoT had become mainstream enough to warrant its own standard (IEC, 2025). The reason this matters is that it gives asset owners a recognised reference for securing exactly the kind of integration brownfield projects involve.
Regulation reinforces the point. A connected industrial gateway is a product with digital elements under the EU Cyber Resilience Act, which brings secure-by-design and vulnerability handling obligations, while ENISA's threat data shows operational technology is an increasingly deliberate target as networks converge (ENISA, 2025). Designing segmentation, authentication and monitoring into the integration from the outset is far cheaper than retrofitting them after an incident.
Takeaway: Connecting legacy equipment removes the safety of isolation, so secure-by-design integration aligned with IEC 62443 and the Cyber Resilience Act must be built in from the start.
The strategic lesson is that brownfield integration rewards restraint. The objective is not maximum connectivity but the right connectivity: capturing the data that drives a clear outcome, such as predictive maintenance or yield optimisation, while leaving stable operations undisturbed.
A measured approach starts with a single high-value use case and the minimum integration needed to serve it, then expands as the data proves its worth. Manufacturers across the Nordics and DACH, where production estates often span many equipment generations, benefit particularly from this incrementalism because it lets them modernise without betting the operation on a single large programme.
The interpretive point is that the same integration architecture that captures value also manages risk. A well-designed edge and gateway layer is the place where data is collected, where protocols are translated, and where security controls are enforced, which makes it the natural foundation for everything that follows.
Takeaway: Brownfield IoT succeeds through incremental, use-case-driven integration that captures high-value data while leaving stable operations and capital largely intact.
Industrial IoT lives or dies at the point where modern systems meet legacy equipment. The value is real and concentrated in the factory, but it is gated by interoperability and security rather than by analytics or cloud capability.
For manufacturers and operators across the EU, the durable approach is additive integration: edge computing and protocol translation to bridge old and new, security designed in rather than bolted on, and a use-case-led rollout that proves value before it scales. Handled that way, decades of installed equipment become an asset for the connected future rather than a barrier to it.
The standard approach is additive rather than replacing equipment. Protocol gateways translate legacy fieldbus or serial data into modern formats, edge computing processes data close to the machine, and non-intrusive data capture reads from equipment without altering its control logic. This lets operators extract value from existing machinery while leaving its core function and reliability untouched.
Edge computing processes and filters data at or near the machine before sending it onward. This reduces latency for time-sensitive decisions, lowers the bandwidth cost of sending everything to the cloud, and keeps sensitive operational data closer to the source. In brownfield environments, the edge layer is also where protocol translation and security controls are commonly enforced.
Connecting previously isolated equipment removes the protection that isolation provided, and each newly connected device becomes a potential entry point with physical consequences if compromised. A connected industrial gateway also falls under the EU Cyber Resilience Act as a product with digital elements. The IEC published dedicated guidance, IEC PAS 62443-1-6, in 2025 for applying industrial security standards to the Industrial Internet of Things.
Usually not. Replacing functioning machinery is rarely justified on connectivity grounds alone, and production cannot be halted for a wholesale rebuild. The more effective approach is incremental, use-case-driven integration that captures high-value data first, then expands as the value is proven, leaving stable operations and capital largely intact.
What is the Internet of Things? – McKinsey – 2024 – https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-the-internet-of-things
IEC PAS 62443-1-6:2025, application of the 62443 series to the Industrial Internet of Things – IEC – 2025 – https://webstore.iec.ch/en/publication/102885
ENISA Threat Landscape 2025 – ENISA – 2025 – https://www.enisa.europa.eu/topics/cyber-threats/threat-landscape
Cyber Resilience Act, summary of the legislative text – European Commission – 2025 – https://digital-strategy.ec.europa.eu/en/policies/cra-summary