Something new is forming on factory floors. AI agents, independent, context-aware and task-oriented, act as a third layer of intelligence. Not a replacement for what came before, but a layer that complements and elevates it. These agents are not limited to a single screen or workflow. They move between systems, interpret context through semantic data, and solve problems across borders.

Think of them as collaborators with domain expertise baked in. They don’t just respond to commands; they interpret objectives from data and instructions. Once narrowed down, they can navigate the data, evaluate outcomes, and coordinate actions. The result is not only a smarter device but a more adaptive factory.

What distinguishes this development is the shift from passively reporting problems to actively solving them. Agents are not there to log information or raise alerts. They act with purpose, looking at tasks and taking action to resolve disruptions before they grow into bigger issues. This evolution changes not only how manufacturing systems work, but how problems are anticipated and managed.

The agent as co-worker

These systems break the logic of traditional software. Most business platforms are fixed structures: interfaces on top of databases, bound by business rules. Agents act differently. They connect to the same data but make decisions based on context. They don’t require a user to click a button; they need a problem to solve.

Basically, they should know what they are talking about. An agent designed for production cannot rely on generic logic. It must understand engineering terms, operational constraints, and supply chain nuances. That’s where domain-specific expertise comes in, combined with semantically organized data.

The power of semantic data becomes apparent in these scenarios, as we see in our own factories, which link voltage spikes, supplier delays, and the drop in yield into a narrative allowing agents to act proactively. They create connections between departments that rarely speak. This can link maintenance data to design documentation or detect recurring defects tied to upstream variables. What once required a team of experts, and a meeting room can now be initiated by a trained agent.

This is not a theoretical promise. It solves a real problem: fragmentation. Most manufacturers still operate in silos, by system, department, or geography. Information does not flow easily. Understanding is lost. Agents offer a way to rebuild that continuity, not by changing the company, but by connecting to its knowledge.

And they don’t just collect data. They did it. For example, the scheduling agent, more than flag conflicts; it can reshuffle shifts, reassign workers, and communicate updates in real time. The emphasis is on initiative, not just alerts.

As these agents take on more responsibility, their role has become that of a digital partner in live factory settings. In many cases, these are multi-agent systems, especially when responsibilities expand. It’s not just lines of code running in the background. They develop working memory, adapt to new scenarios, and respond to consequences. In some cases, they can surpass their human counterparts in consistency or speed. But the goal is not competition, it’s collaboration. Let people focus on strategy and judgment. Allow agents to manage pattern recognition, coordination, and routine interventions.

Human in the loop, by design

Autonomous systems always make the headlines. But on the factory floor, the real goal is reliability. And that means keeping people involved. The most successful agent systems are those that support rather than replace human expertise. They present options, demonstrate their logic, and defer when confidence is low. Operators remain in control, but are better informed. The result is more confidence and better decisions.

The shift can be seen on factory floors today. Some supervisors now coordinate agents and people. Engineers use agents to test hypotheses. Maintenance teams work alongside diagnostic agents who explain what they see and why it’s important. Organizations are beginning to reflect this change. Job roles begin to include responsibility for agent orchestration. The agents themselves are given tasks, benchmarks, and performance reviews.

That opens the door to better accountability. When an agent flags an issue, the chain of reasoning appears. When it makes a recommendation, the source data is clear. This visibility is not a nice-to-have; it is essential. In regulated industries, in safety-critical systems, and anywhere decisions matter, trust depends on transparency.

The cultural shift this implies is not small. For some, it may be the first time a non-human entity has been considered a contributor. This raises new questions about training, management, and ownership. Who reviews the agent’s performance? Who is responsible if they make mistakes? These are not just legal or technical concerns. These are questions about how we create partnerships with machines that are no longer passive tools but active participants.

From use case to intelligence infrastructure

Most of these begin with narrow tasks. Scheduling. Diagnostics. Regulatory checks. These are good test bases: restricted, measurable, and high-impact. But the long-term opportunity goes beyond point solutions.

To create real momentum, manufacturers need to think in terms of platforms. Agents should be modular, composable, and easy to deploy. They should not be locked into any one vendor or one system. Instead, they must sit atop a shared infrastructure that supports semantic data, interoperability, and decentralized execution.

The real challenge, of course, is the current environment. Most plants are a patchwork of legacy systems, vendor-specific formats, and inconsistent standards. Making agents work in that situation requires a new layer of connectivity. That’s where semantic data models come into play. They allow agents to work across systems without rewriting everything underneath.

This opens the door to experimentation. A sustainability agent monitors energy use, flags inefficiencies, and suggests optimizations. A quality agent identifies patterns in defective data and correlates them with upstream variables. A supply chain agent monitors risks and adjusts plans before disruption occurs.

Each starts as a use case. But together, the agents begin to form an ecosystem, often acting as multi-agent systems. And the more they collaborate, sharing data, insights, and context, the more valuable they become. Success at this stage depends on openness. An agent that improves timing in one plant should be able to do the same elsewhere. Portability, scalability, and repeatability will define which models survive. Those built with siled logic or black-box reasoning will struggle to gain traction in large enterprises. Interoperability is no longer a bonus; this is the baseline.

Trust must be earned, not placed

Factories run with precision. When something goes wrong, there are real consequences, downtime, waste, and even safety risks. So, trust in digital systems is not based on innovation. It is based on performance. Trust is now earned by agents who demonstrate accuracy, consistency, and transparency. Their logic is open to inspection, their actions are traceable and their behavior conforms to industry norms, not just technical feasibility.

It’s not just about risk. It’s also about scale. In practice, cost savings and reduced downtime are already measurable. Early deployments show that an AI agent can deliver savings of around €1 million per plant per year. A system that worked once, in a pilot, proves a point. A system that works every day, under pressure, proves its worth. That’s the bar for agent intelligence to create.

And relevance is important. The best agent is not the most complex; it is one who understands the task at hand. That means built with the operator in mind, not just the data scientist. It means solving problems identified by people. When agents help people do their jobs better, they become more effective. Otherwise, they will disappear.

Looking ahead, the manufacturers that lead are not the ones with the best dashboards or the biggest models. They will put intelligence where it matters, in workflow, in decisions, and in relationships between humans and machines.

Factory 2030 is not about firing people. It’s about the reality that’s unfolding on factory floors today: people supported by accountable, transparent digital partners.

The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of luck.