Predictive maintenance gets smarter at spotting equipment faults without failure data

I-care Group is adding an AI system designed to detect equipment anomalies without relying on past failures, after acquiring the assets of French technology company Amiral Technologies.

The Belgian predictive maintenance specialist announced the acquisition on Sept. 17. The deal includes Amiral’s DiagFit technology as well as its research and sales teams.

DiagFit gives I-care another way to analyze complex equipment behavior. It can work with data from multiple sensors and operating conditions, then identify changes that fall outside an asset’s normal operating pattern.

The technology could have particular relevance in industries where equipment failures carry high costs and historical failure data remains limited. I-care said the acquisition expands its ability to serve sectors including defense and nuclear operations.

Spotting faults early

DiagFit does not need a large library of previous failures to identify an anomaly. Instead, it uses historical operating data to establish how a complex asset behaves under normal conditions.

Once that baseline exists, the system can compare incoming operational data against it. A deviation from expected behavior can then trigger an indication that something has changed.

That distinction matters when engineers monitor equipment that rarely fails. A satellite or helicopter, for example, can accumulate extensive operational data without producing enough failure events to build a conventional failure-based model.

DiagFit was developed by Amiral Technologies, a Grenoble-based company that originated as a CNRS spin-off. Its technology has been applied to complex assets including satellites and helicopters.

Fabrice Brion, I-care Group’s founder and CEO, said the acquisition strengthens the company’s ability to process difficult industrial datasets.

He also pointed to the technology’s potential in defense and nuclear environments. Those applications can involve equipment where early detection and controlled data handling carry particular importance.

I-care will progressively integrate DiagFit into I-see, its AI-enhanced predictive maintenance platform. The move will bring the technology’s anomaly detection capabilities into I-care’s wider reliability offering.

Keeping data onsite

DiagFit also supports on-premises deployment, giving customers an alternative to sending operational information into cloud infrastructure.

That capability could become significant when organizations need to keep industrial data inside their own networks. Defense programs and nuclear facilities can face strict security requirements around operational information.

Cloud deployment remains available when customers can use it. The on-premises option instead allows the same predictive maintenance approach to operate within a customer’s existing infrastructure.

The architecture also gives engineers more control over where machine data gets processed. That can remove a potential obstacle when organizations consider AI-based monitoring for sensitive equipment.

Scaling predictive maintenance

The acquisition gives DiagFit access to I-care’s international commercial operation. Amiral’s R&D and sales personnel will join I-care as part of the transaction.

Dogan Kaban, formerly Amiral Technologies’ business development manager, said the combination creates an opportunity to take DiagFit to more industrial customers.

He also highlighted an important feature of the system. DiagFit can identify abnormal equipment behavior without requiring previous examples of failures in its training data.

That approach gives I-care another tool for dealing with machinery where failure events are uncommon but operational data is abundant. The company now plans to fold the technology into I-see progressively.