Maintenance moves from firefighting to connected work

Every factory has a maintenance system that exists outside the software. It sits in notebooks, spreadsheets, toolboxes and the memory of the person who has fixed the same fault for 20 years. That hidden system can keep production moving, but it becomes a liability when factories are short of skilled people, running connected assets and trying to make faster decisions from scattered data.

Jason Afara, director, global high velocity professional services at Rockwell Automation, and Sunil Koticha, director, automation asset management at Rockwell Automation, see maintenance changing because the work now reaches far beyond equipment repair. Afara says the people behind the work are still too often missing from the discussion. “We hear a lot about limiting downtime and the right schedule, and those are true,” he says. “But at the end of the day, it’s really about the people, the electricians, the millwrights, planners who are doing the work, and then even the people who are impacted by the work that’s done.”

The next phase of maintenance is not simply a better way to log jobs. Connected assets are creating signals about performance, condition and risk, while digital tools turn those signals into work orders, technician guidance, parts planning and conversations with production. The value comes from changing the daily rhythm of maintenance so that data reaches the people who can still act on it.

Maintenance cannot remain a response function

Factories have always understood that maintenance matters, but the measurement of maintenance has often been too narrow. Teams were judged on whether they responded quickly, cleared work orders or followed a calendar schedule. Those indicators still have a place, but they can also encourage busy work that does not improve asset health.

Afara is direct about the risk of measuring activity rather than outcome. “If we just focus on completion time, that can lead to pencil whipping or making sure that the records show what needs to be done, but not actually doing the work,” he says. “It’s important to pick the right strategy for the right piece of equipment that you have in your facility, rather than extracting everything from an OEM manual and uploading it and then following the dailies, the weeklies and monthlies.”

Koticha describes the same shift from another direction. Maintenance used to be a calendar activity applied to equipment whether it was needed or not, but connected assets now provide a different starting point. “What has really changed is the assets can now tell us how they’re doing,” he says. “There’s a heartbeat out in all the assets that are available. How do we leverage that data and insights that our assets are actually talking to us about, and drive that into our maintenance workflows so we can drive productivity, resiliency, asset utilisation holistically?”

The difficult part is not only technical. Koticha argues that data-driven maintenance depends on change management because decisions have to become part of the plant’s operating cadence. If a team begins the day by discussing what the data is showing and what actions need to follow, maintenance becomes part of production planning rather than an interruption to it.

The workflow must connect

Traditional maintenance is fragmented because different teams own different pieces of the problem. Controls, reliability, cybersecurity, purchasing, planning and operations may all hold information that matters, but the work breaks down when those teams cannot see the same picture. Koticha says the challenge is to unify people around the work that needs to be done, particularly in factories with a mix of connected and disconnected assets.

Afara sees workflow connection as a way to make maintenance more thoughtful. If a machine has run for 1,000 or 3,000 hours, that information should influence the level of work performed, the parts required and the timing of intervention. Without that connection, plants can over-maintain equipment, repeat tasks that no longer make sense and frustrate the technicians expected to carry them out.

“Creating that connection of data flow allows us to have that right strategy and the right level of work, and we’re not over-PMing, which can lead to low morale or a disconnected workforce,” Afara says. “If I’m coming in and doing the same work each day, but we’re not seeing a change of results, I’m not going to align with the strategy of that maintenance manager.”

DLG Group shows the operational value of creating a shared maintenance view. The Danish agricultural feedstock and vegetable oil producer had multiple sites with similar processes but different machines by make, model and age. Without a common inventory or knowledge base, technicians often had to resolve recurring problems from first principles, while spare parts could be duplicated at one site and unavailable at another.

By deploying Fiix CMMS across nine locations, DLG created a central source of truth for maintenance issues, spare parts and known fixes. Within a month, the initial rollout had already improved mean time between failure, and the company targeted a 10 percent improvement in uptime and maintenance efficiency within a year. The same issue could now be understood across sites and roles rather than rediscovered locally each time.

Better data changes the role of people

The maintenance manager’s job is also changing. Afara says the role was historically occupied by the best person at fixing equipment, someone who could teach others how to complete corrective work. That practical knowledge still matters, but the same leader is now expected to understand analytics, contribute to capital planning, interpret AI insights and work with other departments to secure production time before failure occurs.

That creates a new kind of pressure. Maintenance teams are dealing with skilled labor shortages and thinner engineering coverage, while newer technicians need access to knowledge that experienced people often carry in their heads. Mobile access, searchable documents and AI assistants can help, but only if the underlying information is captured and trusted.

Afara points to the ability to extract knowledge from manuals, standard operating procedures and troubleshooting guides, then make it available to technicians through AI tools. “That can help with onboarding. That can help with confidence. That can help with communication,” he says. “The purpose is not to replace experienced technicians, but to make hard-won knowledge available beyond one shift, one site or one person.”

Dunlop Aircraft Tyres provides a practical example of how quickly maintenance changes when history, scheduling and reporting are brought into one system. Before adopting Fiix CMMS, the UK manufacturer relied on paper, Excel and an in-house Microsoft Access database that could not schedule or plan maintenance work. After implementation, the team reported improvements within the first few months, automated daily reports for plant managers and increased mean time between failure in the curing department by 21 per cent between February 2023 and July 2024.

That movement from record keeping to decision support becomes more powerful when it changes how maintenance teams prioritise the work itself. Nói Síríus, the Icelandic chocolatier, had production-line systems that were not being monitored, leaving staff reacting to urgent call-outs with little reliable data on likely failures. After implementing Fiix CMMS across crucial production-line machines, the company consolidated its job queue, reduced the ticket backlog and moved toward preventive maintenance. Within six months, Rockwell Automation and Nói Síríus began work to network the entire production facility.

Maintenance becomes a shared responsibility

Predictive maintenance can sound like a technology project, but its value only appears when someone acts. Koticha says trust is essential because plant teams will not act on insights that create noise or waste their time. The same trust runs in both directions: the system needs accurate work-order information from people, while people need confidence that the recommendations coming back are relevant.

“When we build that trust with the right insights and the right actions are taken, how do we close the loop and ensure that we leverage that insight and data that whoever’s worked on that asset, we have captured it, so we can learn from it the next time?” Koticha says. “The information that you have captured in your work order is extremely critical, so that we can continue to learn and provide better context and better insights as time progresses.”

That closed loop also changes the relationship between maintenance and operations. A maintenance leader asking for a line during the working week is more credible when the request is supported by asset history, risk information, parts availability and a clear explanation of the consequence of delay. Maintenance then becomes less like a service called after something breaks and more like a contributor to production reliability, safety and quality.

Koticha draws an analogy with safety, where responsibility is no longer confined to a single function. “Safety applies to everyone,” he says. “I think in that same context, though the activity may be done specifically by the maintenance team, maintenance is the name of the game for everyone.”

That is the real shift behind connected maintenance. The future is not simply fewer failures, better dashboards or more predictive alerts. It is a shared operating model in which asset data, work orders, skills, parts and production priorities are connected tightly enough for manufacturers to act before maintenance becomes downtime.