Advanced asset intelligence has people in the loop: MOVUS boss

Effective connection of advanced technology and human experience at critical points in fast-moving asset management cycles is a key reason new-generation prescriptive plant maintenance systems such as MOVUS’ PlantOS are succeeding where predecessors failed, CEO Sanjeev Kumar said this week on a webinar.

Kumar spoke with metals producer Nyrstar’s Samitha Wijesinghe, an experienced reliability engineer, about the value of AI-driven machine health analytics in a new era of heavy industry maintenance.

But the MOVUS boss started by highlighting a shift in predictive maintenance that has seen technology adoption progress.

MOVUS, an Australian 2015 start-up acquired last year by Infinite Uptime, was among a host of new companies that brought predictive maintenance platforms to market. “You will see 99% of [the firms] have died,” Kumar said. Most had disappeared in an avalanche of plant maintenance alarms and alerts that “created big-time pushback from users”.

“There were so many alerts the user said, should I focus on the production side or should I focus on alert management,” Kumar said.

“We are one of the 1% of companies who survived and moved beyond [early] predictive maintenance.

“AI came into the picture, internet came into the picture, IoT sensors become more efficient and effective, and the change from the traditional approach to a modern and prescriptive approach [meant] detection became more of a diagnostic and the alerts, which used to come earlier, were converted into recommendations.

“A fancy dashboard doesn’t help to run the plant.

“But a guided intelligence system is helping the maintenance and the planning and the operation team to take an informed decision.

“This is where we moved from predictive maintenance to prescriptive, where there is a human in the loop.

“The right people can act at the right point where the need [is real].”

MOVUS launched its PlantOS prescriptive AI platform to the mining and metals market at the 2025 International Mining and Resources Conference in Sydney, Australia, where it also picked up the inaugural IMARC 2025 Mining Beacon Breakthrough Innovation Award for Ultra vSense, a world-first piezoelectric sensor combining vibration, temperature and RPM measurement in a single, rugged device.

Kumar said this week the company’s flagship platform had been deployed at more than 900 plants worldwide, including mines, cement, steel and other metal-making plants.

“What we observe wherever we have deployed a solution is that we have been able to reduce unplanned shutdowns by over 90%,” he said.

“And not only that, the feedback is the plant maintenance planning has improved by 20-to-30%.

“We sit in the middle between the OT [operations technology] and IT [information technology] layer and we ensure that the IT layer is getting the right intelligence from the OT layer. [PlantOS] works as a standalone platform. We provide data through API so the same platform can integrate with any other system to bring operational and production data into the same decision-making environment.”

MOVUS says seamless integration with existing PLC, SCADA and distributed control systems, capturing all critical process parameters, creates the context that enables effective prioritisation and action by maintenance teams. Data is securely transmitted to the collaborative AI within the PlantOS ecosystem for analysis and optimisation.

“The whole idea of the [PlantOS] system is that we should be able to reduce unplanned downtime, we help you to optimise your energy consumption, reduce your process deviation, etc,” Kumar said.

Nyrstar’s Wijesinghe, part of a condition monitoring team supporting critical assets at the compan’'s lead smelting and refining operation at Port Pirie in South Australia, said the company’s transition to a full diagnostic prescriptive AI platform in the form of PlantOS was in its early stages but showing positive signs.

“We are confident that it’s supporting us to save a great deal of time,” he said.

On the specific topic of machine health alerts generated previously by MOVUS’ MachineCloud predictive maintenance platform, versus the PlantOS prescriptive system, Wijesinghe indicated the contrast was significant.

“I can recall one example out of many,” he said.

“We have one exhaust fan where we got about 30 alerts in a week with MachineCloud and now it has been reduced to one single prescriptive analytical report on which we acted.

“We saved a catastrophic failure.”