This article is intentionally forward-looking. It represents a view of how industrial digital services, AI agents and business-to-business interactions may evolve over the coming decade rather than a definitive prediction or formal strategy.
While many of the underlying technology trends discussed are already emerging across industry, the pace of adoption, standards development, customer behaviour and commercial models will vary by sector, company and geography. The scenarios described should therefore be considered as informed possibilities designed to stimulate discussion, challenge assumptions and help organisations think about how they may position themselves for an increasingly agent-enabled future.
The purpose is to ask a strategic question: what could change if specialist industrial knowledge became as accessible to customer systems as data is today?
The next industrial revolution will be built on relationships
For more than a century, industrial companies have competed through product quality, engineering capability and service. Digital transformation added a new dimension: the ability to collect, share and analyse data.
Today, most industrial suppliers have embraced this digital era. They offer connected assets, analytics platforms and APIs, which allows customers to monitor performance remotely, identify trends and integrate operational information into enterprise systems.
Yet despite significant investment, a fundamental challenge remains.
Customers do not wake up wanting data
They want outcomes.
They want lower energy consumption, fewer breakdowns, reduced maintenance costs, improved uptime, greater sustainability and better operational decisions.
For decades, the gap between data and outcomes has been bridged by people. Engineers, consultants, reliability specialists and subject matter experts have interpreted information and translated it into action.
But that model is now beginning to change.
Just as APIs transformed the exchange of data between software systems, agent-to-agent communication could transform the exchange of knowledge between organisations. Suppliers that understand this shift may become more closely embedded in their customers’ decision-making. Those that do not risk becoming providers of raw data within someone else’s ecosystem.
The end of the dashboard era
During the first phase of industrial digitalisation, the objective was connectivity.
-
Connect the asset.
-
Collect telemetry.
-
Visualise performance.
-
Generate alerts.
This remains important, but it is no longer enough to differentiate one supplier from another.
A connected steam trap, pump, heat exchanger or boiler generates data. An analytics platform then converts that data into insights. But a person must still interpret the issue, assess its significance and decide what to do.
In practice, this creates friction.
-
An alert appears.
-
An engineer investigates.
-
A maintenance planner creates a work order.
-
A supervisor approves the action.
-
A technician is scheduled.
The process works, but much of the effort occurs after the insight has been generated.
The next wave of transformation focuses on what happens after insight.
Instead of merely identifying a problem, digital systems will increasingly understand context, recommend actions, orchestrate workflows and initiate outcomes. In that sense, the direction of travel also reflects Industry 5.0: using advanced technology to support more human-centric, sustainable and resilient industrial decisions.
The future is not about providing customers with more information.
It is about providing customers with expertise on demand.
The emergence of the industrial agent ecosystem
Technology providers are already building the foundations for agent-to-agent (A2A) communication and emerging protocols are being designed specifically to allow autonomous agents from different organisations to discover one another, exchange information, coordinate activities and collaborate securely across organisational boundaries.
Consider a future manufacturing facility.
- One agent monitors performance across multiple sites.
- Another manages work orders.
- Other agents support procurement, inventory or energy optimisation.
- When an issue emerges, these agents do not simply display alerts for human review.
Instead, they engage directly with specialist supplier agents
A customer maintenance agent might ask:
"Which steam assets currently present the greatest financial, or operational risk this month and what action should we priortise?"
A supplier agent could analyse the available asset history, operational conditions, and failure patterns before returning with a priortised recommendation.
The opportunity is significant, but trust will determine whether it becomes operational reality.
A new industrial value chain
One of the clearest lessons emerging from recent discussions with major digital transformation partners is that industrial value increasingly extends far beyond sensing and monitoring.
The traditional digital stack can be viewed in six layers:
Layer 0: Physical Assets
Industrial equipment, sensors and instrumentation generate operational data. This remains the foundation of the value chain.
Layer 1: Connectivity and Data Acquisition
Control systems, industrial gateways, MQTT brokers, historians and edge infrastructure capture and normalise operational information.
Layer 2: Data Ingestion
Secure cloud infrastructure transports, stores and manages information streams from connected sites.
Layer 3: Analytics and Applications
Domain-specific algorithms convert telemetry into actionable information through analytics, condition monitoring, optimisation models and digital twin foundations.
Layer 4: Enterprise Integration
The real transformation begins when insights flow into wider business systems such as ERP, MES, CMMS, maintenance platforms and enterprise workflows. Here, asset intelligence becomes embedded within operational processes rather than existing in isolation.
Layer 5: Intelligence and Autonomy
This is where the future resides.
AI agents, copilots, prescriptive analytics and workflow automation combine domain expertise with enterprise context to drive decisions and actions. Instead of simply presenting information, systems begin to orchestrate outcomes.
This progression represents a shift from monitoring assets to managing decisions.
The competitive battleground is moving up the stack
Historically, industrial companies competed primarily at Layers 0-3.
Who had the best hardware? Who collected the most accurate data? Who delivered the best analytics?
Those questions remain relevant.
But future differentiation increasingly occurs at Layers 4 and 5.
Customers are already demanding:
- Enterprise integration.
- Automated maintenance workflows.
- Cross-system visibility.
- AI-driven decision support.
- Multi-site optimisation.
- Conversational interfaces.
- Autonomous recommendations.
Recent industry collaboration frameworks show a common theme: customers want operational intelligence embedded directly into maintenance systems, production environments, enterprise platforms and decision processes. They want asset insights converted into business outcomes.
For suppliers, the opportunity may not be limited to providing another sensor or dashboard. It may lie in becoming a trusted and accessible source of expertise across the customer organisation.
The rise of "Expertise-as-a-service"
This development could create a new commerical catgegory. In this model, suppliers provide:
- Asset intelligence.
- Engineering knowledge.
- Operational recommendations.
- Decision support.
- Continuous optimisation.
The customer no longer purchases a report. They subscribe to a continually improving digital expert. Human specialists remain critical, but their role evolves. Rather than repeatedly answering the same questions, they train, govern and enhance the expertise embodied within digital agents. Expert knowledge becomes infinitely scalable.
The important shift isn’t just from dashboards to AI, but from periodically delivering information to making our engineering expertise digitally accessible at the point a customer needs to make a decision.
Our advantage is not simply having data. It is the combination of decades of steam expertise, application knowledge, observations and optimisation experience that could increasingly be turned into trusted digital intelligence.
What this means for companies like Spirax Sarco
For specialist industrial organisations, this shift creates an extraordinary opportunity.
Many businesses possess something far more valuable than data. They possess expertise. Decades of engineering knowledge. Millions of operational observations. Thousands of customer engagements. Deep understanding of process optimisation, reliability and performance improvement. Historically, much of that expertise has been delivered through people.
The next stage of digital transformation is about making that knowledge available in a form that customer systems, customer agents and customer workflows can consume directly.
The future relationship between suppliers and customers
The future industrial relationship may look very different from today's.
Customers will have their own agents. Suppliers will have their own agents. Enterprise orchestrators will coordinate interactions between them. The most successful suppliers may become those whose knowledge is most easily discoverable, consumable and actionable within those ecosystems.
In that world the product is not the sensor
The product is not the dashboard.
The product is not even the data.
The product is the expertise behind the recommendation and how it is delivered to the user.
And when expertise can flow directly between organisations through trusted AI agents, the relationship between suppliers and customers transforms from periodic engagement into continuous collaboration.
That may prove to be the most significant industrial business model shift since the arrival of the internet itself.
Authored by Tom Johnson, Head of Commercial Offers - Digital & Services at Spirax Sarco