ZYFRA displays Ai predictive maintenance software at EMO

The software is a part of ZYFRA’s MDCplus system and predicts remaining useful tool life. The solution helps to reduce unplanned downtime by 10% by detecting anomalies in technological processes and predicting equipment failures. ZYFRA is among the pioneers offering a PM solution to the market which, unlike other solutions of its kind, does not require installation of any additional sensors.

“Downtime is very costly, as failure of functional equipment can lead to productivity and efficiency losses. The PM solution helps to identify defects or predict failures such as a worn spindle, thereby helping eliminate faulty products and avoid financial losses,” said Ilkka Saarinen, Sales Director of Nordic markets at ZYFRA.

The software is based on AI technology. ML models use data from the CNC, find correlations between different parameters and give a prediction of tool life. The key aspect of this solution is data. It has to be historical, so the algorithms can see the whole tool lifecycle, and also complete, meaning it should contain crucial parameters (cutting parameters, spindle load, spindle and feed speed as a minimum). For example, based on historical data for one month, the model can give a prediction of tool life on a 15-minute horizon with 85% prediction accuracy.

The essential platform for collecting and storing this data is ZYFRA’s IIoT-based MDCplus, which is a real-time machine monitoring and manufacturing data collection system. The advantage of the solution is that it measures not only superficial operational parameters like temperature or power consumption, but also deeper ones such as spindle rotation rate within a specified work shift timeframe.

At EMO, Andrey Lovygin, ZYRA’s International Business Development Director, presented the Predictive Maintenance module at the New Technologies – Future Opportunities forum. In his presentation, he talked about Industrial AI as a development of Industrial IoT ideology and explained how it can provide added value for manufacturing companies.

“IoT Analytics predicts that the Predictive Maintenance market will grow from USD 1.49 bn in 2016 to USD 10.9 bn in 2022. At present, however, Predictive Maintenance solutions do not have a significant presence in discrete manufacturing. While holding a leading position in IIoT implementation (according to IoT Analytics, spending on IIoT platforms for discrete manufacturing will account for 53% of the total), discrete manufacturing is lagging behind in terms of AI application. AI based solutions in the sector are more like research projects than working products. ZYFRA has connected over 8,000 machine tools. As a result of our work, with all the data now accumulated, we are able to offer the market our new Predictive Maintenance solution, “said Andrey Lovygin.

At EMO Hannover, ZYFRA also officially announced its expansion into Turkey, having signed an agreement with TANDEN TAKIM TEZGAHLARI, a large Turkish distributor of CNC machines. This year, the company has already entered the markets of Peru, Chile, Malaysia and France, and is currently operating in over 10 countries around the world.

ZYFRA was founded in November 2017. The company develops and invests in industrial digitalization technologies, and is improving the Industrial Internet of Things and Artificial Intelligence environment. ZYFRA’s key target sectors are the machinery, metallurgy, mining, oil & gas, and chemical industries. ZYFRA promotes readymade industry solutions in predictive analytics and data analysis, tech process optimization, and machinery and floor staff monitoring.